NSAID-INDUCED GASTRIC ULCERS

Prostaglandins and rnucosal defensive mechanisms

GERALD P MORRIS, PHO, TODD E WILLIAMSON, MSC, T AIMI T HYNNA, MSC

ABSTRACT: The first line of mucosa! defence includes the juxtamucosal unstirred layer/pl I gradient and the apical surface~ of the luminal epithelial cells. Many damaging agents, including nonsteroidnl anti-inflammatory drugs (NSAIDs), can overwhelm Lhese defences and destroy extensive regions of the luminal cpiLhelium. This damage is readily Lolerated in the normal mucosa. Furthermore, a combination of increased mucosa! bloodflow, epi thelial migra­tion, mucus release, and efflux of bicarbonate- rich fluid usually allows rapid recovery of mucosa I integrity. In the presence of vascular damage and congestion, however, luminal acid can kill mucosa! cells and destroy the substrate necessary for repair (by epithelial migration ). Damage of this type results in the production of hemorrhagic erosions, which may then develop into chronic ulceroinflam­matory disease if healing is prevented by exec&, lu minal ac id or by impaired mucosa! immune response. Endogenous and exogenous prostaglandins could affect all aspects of the mucosa! defensive response~, from the juxcamucosal unstirred layer/pH gradient ( via effects on secretion of bicarbonate, acid and mucus, as well as stimulation of fluid efflux) to the function of the mucosa! immune system. Protection against the acute damage produced by topically administered NSAIDs or concentrated ethanol can result from either administra­tion of prostaglandins or topical applicaLion of'mild irritants'. This is referred to

as 'adaptive cytopr~tection'. Parenterally adm in istereJ NSAIDs can also produce mucosa! erosions. Proteclion against this type of damage may depend on the effects of proscaglandins on neural and contractile elements in the mucosa. Studies on animal models also suggest that by preventing acute hemorrhagic erosions, prostaglandins may prevent the development of chronic ulcer in sus­ceptible individuals. Can J Gastroenterol 1990;4(3 ):95 -107

Key Words: Adaptive cytOf)rotection, Cytof)rocection , Gastric ulcer, Microvascu.la­ru.re, Mucus-bicarbonate barrier, Prosraglandin, Unstirred water layer

Les prostaglandines et les mecanismes de defense de la muqueuse

RESUME: La premiere ligne de defense de l'estomac est le gradient de pH de la couche juxtamuqueu~e non perturbce et la surface externe des ce llules cpithclia les. De nombreux agents nocifs, parmi lesquels les anti-inflammatoi res non steroi"diens (AINS), reuvent aneantir ces facteurs de protection et detruire

Gascromcestmal Disease Research Unir and De/>artment of Biology. Queen's University, Km11sw11 , Onumo

Cor,·es/>ondence and re/mncs: Dr GP Mon-is, De/Jartmenr of BioloRY. Queen's University, Km11swn , Ontario K7L3N6. Telephone ( 613) 545-6 130. Fax (61 3) 545-6617

CAN J GASTROENTEROL VOL 4 N() 3 M AY 1990

SUPPRESSION OFACID SECRETION BY

H 2 recer ror antagonists remains the basi~ for current therapy of rerric ulcer. The most o bvious and well defined therapeutic value of rro~raglan­dins in the treatment of pcpuc ulcer results fro m rheir anrbecretory proper­ttcs ( 1 ). By decreasing the quantity and concenlration of secreled acid, prosta­glandins can promote healing of peptic ulcer and min1mi:c the pmducrion of acute damage produced hy agents such as aspi rin. When viewed solely m terms of abi li ty co surprcss ac id secretion, prostaglandim do nor offer significant therapeuttc advantages over the Hz recepror antagonises. h is becoming ar­parenc, however, that prostaglandins have ocher potentially benefic ial effects on the mucosa, independent of their effects on acid secretion, which may justify the ir use in cem11n clinical set­tmgs as alternatives to other therapies. Ln particular, prostaglandin have been shown to affect the funclion of several aspects of mucosa! defence.

G ast ro intestinal mucosal defences are dynamic rather than slatic barriers. Succe sive levels of defences must be overcome during the transition from a broken mucosa I barrier to acu te erosion and finall y chronic ulceroinflammatory d isease (Figure 1). By defining these defences and identifying the agents and mediators which overcome or 11npair them , it is possible lo identify sites and stages a t wh ich lhe mucosa is vu l­nerable and a t which therapeutic inter-

95

MORRISe1al

des regions importances du revetement epithelial. Ce dommage est fac ilement tolere clans des conditions normales. De plus, !'augmentation de l'apport sanguin au niveau des muqueuses, la migraLion epithelia le, la production de mucus e t l'afflux d'un liquide richc en bicarbonates conjugucnL leur ac tio n pour permett re habituellement a la muqueuse de retrouvcr rapidement son integrite. En presence de lesion vasculaire et de congestion, neanmoins, l'acide gastrique peut derruire les cellules de la muqueuse et le substrat necessa ire a la regeneration ~pithelialc. Si l'exces d'acide gastrique ou une defaillance de la reponse immunitairc de la muqueuse empeche la c icatrisation, ii se produit des erosions hemorragiques qui peuvent ensuite resulter en une affection ulcero-inflammatoire. Les prosraglan­dines endogenes et exogenes pourraient bien affecter tous les aspects des rcponses defensives de la muqueuse, a commencer par le gradient de pH de la couche juxtamuqueuse non perturbee (par leurs activites sur la secretion de bicarbonates, d'acide et de mucus, ainsi que la stimulatio n de l'afflux liquidc) et jusqu 'a la fonction du systeme immunitaire de la muqueuse. La prevention de tout dom­mage grave provoque par les AJNS ou l'a lcool concentre peut s'obtenir en administrant soil des proscaglandines soit des 'irritan ts doux' - c'est a dire par 'cywprorection adaptative.' Les AINS pris par voie parenterale peuvem aussi entrainer l'erosion de la muqueuse. La protection peut dans cc cas dependre des effets des prostaglandines sur !es elements nerveux et con t racciles de la muqueuse. Certaines etudes chez !'animal suggeren t egalement qu'en prevenant les e rosions hemo rrag iques a igues, lcs p rostagland in cs e mpec hent pe ut-etre le developpement de l'ulccrc chronique chez les sujets sensibles.

BROKEN MUCOSAL BARRIER

unstirred water layer 0 C)

epithelial resti tution

@ surfactants

0 extrusive cell loss

0

release of gel mucus

ACUTE EROSION

bicarbonate secretion mucosal blood flow

.. r •,.,

acute inflammation mucosal immunologic barrier

CHRONIC ULCER

accelerated c e ll turnover I healing from ulcer margin

Figure 1) Diagrammauc re(JT'eselltation of gastric mucosa! defences and 1he types of damage which may result when successive layers of defences are overcome. The locations uf various defence.1 on the diagram (co1Tesponding to sites in or external w the surface ep1chelium and within the mucosa, submucosa and muscularis) indicate the aJJproxnnate depth of damage as each successive ,et of defences is overcome. The components of each defensive response (such as the nmnunologic barrier or accelerated cell turnover) can occur throughout the de pth of che )!astric wall

venrion could prevent progression from readily repaired superficial damage to ch ronic ulcer.

Prostagland ins have Lrophic and im­muno mod ulaw ry effects. T hey can : stimu late secre tion of mucus and bicar­bonate; increase mucosa! bloodflow; af­fec t clectrogenic ion t ransport; and bo th stimulate and inhibiL gastroimes­t inal motil ity (2). Prostaglan<lins are also produced in concert with many other med iaLors of inflammation such as leukotricnes and plaLclcL activa ting factor, and may affect their production or actions.

Thus, proscaglan<lins can affect al l stages in the progression of injury and may not o n ly prev cnL the init ial devclopmcm of acute injury, but a lso aid in healing and preventing relapse.

PROST AGLANDINS AND CYTOPROTECTION

The demonstration that prostaglan­dins have both anrisecrctory and ant i­ulcer acLivities in animal models ( l ,3 ), and that certain synthe Lic prostaglan­d ins a rc much mo re po le nt an t i­secre to ry agen ls Lha n na t ural prostagland ins ( 4 ), suggested that these agents may be of Lherapcutic va lue in the reduction of gasLric acid secre tion. A subsequent sLUdy demonstrated thal 'cytoproLective' pro~tagland ins could also prevcnl the development of acute hemorrhagic erosions produced in the gastric mucosa hy necrotizing agents such a~ conce mrated ethanol and strong acids (5). T his study provide<l fu rt he r suppo rt fo r a pole n t ia l therapeutic ro le of Lhese co mpounds in the trea tmcm of ulcero inflammatory d iseases of. the upper gastroin testinal tract. S ince the cytoprotectivc proper­t ies of prostaglan<l ins were manifested a t doses below those which were anti­sec re t o ry, it seem ed that a new therapeutic approach was feasible.

lt was also apparcn l Lhat many agents o ther than prosraglandins were cytoprotective. When applied LO the gast ric mucosa prior to necrotizing agents, so-called 'mild irritants' were shown to produce 'cyropro tccLio n ' equ ivalent to char provided by exo­genous prostaglandins (6 ). This gave rise to rhe concept of 'adaptive cyto-

96 CAN J GASTROENTEROL VOL 4 Nu 3 MAY l 990

Prostaglandins and mucosa! defences

Figure 2) Light microgra/Jh of rat fundic mucosa, 5 mins after exposure to 9% ethanol comaining 16 mM acetylsalicylic acid. lmerfoveolar epirhelial cells are irreversibly damaged and are being slied from the mucosa. (Plastic section stained with wluidine blue x540)

Figure 3) Light micrograph of a biopsy of human antral mucosa IO mins after exposure to 40% ethanol. Ar1'0ws indicate congested subepirhelial collectingven11les. Ultrastrnctural study indicated that most of the inier­foveola,· cells were i1Te,1ersibly damaged. A local, adherem mucus coat (M) has been released from che damaged epithelial cells. ( Plastic section stained with wlwdine blue x350)

protection'. Adaptive cytoprotection was initially found to be abolished by the pretreatment of animals with in­hibi tors of cyclo-oxygenase activity. This was interpreted as evidence that protect ion depended on irritant­induced stimulation of prostaglamlin synthesis. Elevated prostaglandin syn­thesis has also been invoked as an ex­planation for the cytoprotective actions of agents such as sucralfate and the aluminum-containing antacids (7,8).

It has proven difficult, however, to prov ide convinc ing explanations for the mechanisms which underlie the phenomenon of cy toprotcction. Cytoprotection by prostaglandins in the gastric mucosa is clearly not de­pendent on the preservation of the in­tegrity of the surface epitheli um in the face of assault by barrier breakers, but rather results from the prevention of microvasculardisruption and the main­tenance of gastric mucosa! bloodflow (9). In this sense, most experiments have demonstrated indirect cytoprotec­tion where preservat ion of mucosa! in­tegrity is a consequence of effects on the mucosa! microvasculature. Further­more, adaptive cytoprotect ion can occur even when prostaglandin syn­thesis is significantly inhibited by prior administration of indomethacin (10-12). It is a lso necessary to consider the means by which agents othe r than prostaglandins can prevent rnicrovas­cular disruption.

There is evidence that prostaglan­c.lins exert limited direct cycoprotection on in vivo gastric mucosa! preparatiom and gastric cells which are isolated or maintained in cell cu lture (13,14). Such protect inn has heen demonstrated against nonsteroidal anti-inflammatory drugs (NSA!Ds), ethanol and rnuro­cholate, and could play a role in the su rviva l of certain populations of mucosa! cells fo llowing exposure to

damaging Clmcentrarinns of these necrotizing agents (13-1 5). Studies on in vivo systems have demonstrated that direct cytoprotection of the lurninal epithelium of the stomach against con­centrated ethanol can be achieved hy prior exposure co a mild irritant ( 11) or to sucralfate (12). This may be due to a 'histodilutional' effect ( 16), whereby the presence of an irritant-induced fluid efflux and edema, or a protective cover­ing of sucra lfate and mucus, attenuates the concentration of ethanol which comes into contact initia lly with the mucosa. There is no reason to believe that the luminal epithe lium is rendered resistant to concentrated ethanol.

PROST A GLAND INS AND REPAIR OF SUPERFICIAL

DAMAGE BY RESTITUTION Superficial damage to the stomach ,

of the type readily produced by barrier breakers such as acidified aspirin, sodium taurocholate, and concentrated ethanol may result in destruction of

CAN J 0ASTROENTEROL VOL 4 No J MAY l 990

most of the lumina l epithelium (Figures 2.3). Although this damage is irre­versible ar the cellular level, it is often restricted to luminal epithelial cells. If gastric mucosa! microcirculation is not impaired, this type of damage is readily repaired or restored hy cell migration from gastric glands ( 17) . The restitu­tion process is rapid and typically com­plete in about I h in both the stomach nnd small intestine.

Superficial injury is usually fo llowed hy increased gastric mucosa! bloodflow which, in comhinarion with a cap of exfnliatcd cells and mucus, provides a localized alkaline rnicroenvironment within which the influxing aci<l can be neu tralized ( 17 ). Although it may be widespread, superficial damage to the epithelium must be considered a nor­mal event and does not usually produce ulceration. Such injury is a result of the ready entry into cells of the undis­sociated lipoph ilic sal icyla te molecule, the detergent effects of bile salts, or osmotic damage in the case of hyper­tonic solutions. In the case of acetyl­salicy lic acid (ASA), once absorbed into the cells of the gastric epithelium, the molecule encounters a near neutral pH and reverts to the ionized form, resulting in disruption of the intracel­lular ionic balance and local ized ac­cumulation of acid.

ASA also produces extensive and rapid damage to the intest ine within minutes. ln one study on humans, a

97

MORl\hc?ral

Figure 4) Scanningelemon micrograph of cor­rosion cast of rat gastnc mucosa showing sub­mucosal arcery ( A) and terminal arteriole (TA) ar the base of the mucosa. The tenninal arteriole fonns f)rinwry ( I), secondary ( 2) and terciary ( 3 ) lmmches which give rise to rhe mucosa/ caprllanes (x245)

J ose of ASA equivalent to two table ts in a glass of water damaged 25 to 30% of jeJunal villi wi thin 5 min · ( 18). Similar levels of damage have hcen produced in the proximal Jejunum fol-1 owing exposu re to in trn luminal ethanol at peak concentrations of abt1uc 6% (19 ). As in the stomach, chis type of damage was most frequent in the o ldest ce lls (at tht: villus tip~) and ac­companied by blistering and rupture. Ivey et al ( 18) also found a 20% in ­cidence of duodenal lesions I y endo­scopy in rheumatic disease patients receiving ch ronic ASA therapy. T hey n oted that altho ugh ente ric-coated ASA s ignificantly reduced gas tric mucosa! damage, the in c idence of duodenal erosions remained at about 20%.

S uccessful repair by restitution depends on either an absence oflum inal acid or an intact mucosa! microvascular supply which maintains a surply of bicarbonate-rich fluid to the overlying layer of damaged cells and released mucus. In the rresence of vascular con­gestion, acid not on ly kills cells but also destroys the basal lamina (basement membrane), which is a substrate neces­sary for cell migration ( L 7). Following

98

Figure 5) Scanning electron micrograph of ccnTosron cast of wt fundic m11cosa A mbmucos,1l artery ( A) gives rise w arceriob (one of which 11 indicated by an arrow), which prod rice capillarres (CAP) The capillaries conn~ct wiih th t.' t'lJllccun,g 1•enrdes (C\') near the mucosa/ rnrface. Col/ecrrngt·enules drain into a wnou~ plexus /1·/J) wluch lies w the bme of the m11cosa The t•ein (v) penetrates rhe m11srnlarn mHcosa and drarm the 1·enor1.1 plexus IX I 06)

exposure to prnsraglandins, h,w,cvcr, the mucosa i~ typically resistant ni vas­c ular damage and rapid repair ca n proceed. It is sr ill no t clear precbely how th e administration of pro~rnglan­dins or mild irritants prevents necrotiz­i n g agents from causing vasc ular damage. It is clear, howeve r, that the end re sult is the preserv;nion of microvascular integr ity.

ULTRASTRUCTURE OF GASTRIC MUCOSAL

MICROCIRCULA TION The most revealing method for

visualizing the a rchitec ture l)f the gastrointestinal microc ircularion is that of scanning e lectron microscopy (SEM) of corrosion casts. With this technique, the microvasculature is perfused with saline followed by low viscosity plastic. The plastic is a llowed ro pol ymerize, the tissue d igested, and the resulting casts viewed by SEM. The micrnvascular

architecture ofhumnn and mt stomachs is ha~1cally the same (20,21 ). At the mucosa! hase, prnnary, secondary, aml ceruary branches of mucosa! arterioles give nse ro an a~ccnding carillary ar­cade (Figures 4, 5 ). The capillary arcade connect~ with the cnllec ting venulcs rn the uprer regions ot the mucosa (Figure 5 ). There a rc no la teral interconnec­llnns between the ascending capi llaries and the collecting venules. However, the interconnections hetween the capillaries and the multiple connec­tions of tht.: carillaries w1rh the collecr­ing venules result 1n cons1de rnble redundancy (Figure 6).

An additional level of redundancy b introduceJ by the formation of large branches by many of the collecting venules. In the rat fundus, the au thors found that a lmost half (44.8±3.4%) nf the collecting vcnules were formed by fusion of majo r branches (Figure 7 ). More than 95% of these major branches

CAN j GASTROENTEROL VOL 4 N o 3 MAY 1990

Figure 6) Scanning electron micrograph of luminal surface of corrosion cast of rat fundic mucosa. This view of the mucosa/ surface shows the honeycomb arrangement of the subepichelial capillimes and the intercon­nections becween capillaries and a branched colleccmg venule. The four arrows indicate the major branches of the collecting venule and the direction of bloodjl()w (XI 70)

Prostaglandins and mucosa! defences

Figu re 7) Scannmg electron micrograph showmg branching of collecting venules and venous /}lexus ( 11p) from rat fundic mucosa. This is a partial retTugrade cast which u•as produced by infw,ion of plastic through che 111per10r mesencerrc vem. Perfusion u•as stopped before the capillary bed WCI.I filled (xi 00)

converged in t he upper two-thirds of the mucosa. The collect ing venules drain into an anastomosing collecting venule plexus whic h lies at the base of the mucosa, adjace n t to the muscularis

mucosa (Figures 7 ,8 ). The mucosa! mic roc irc ul a tion ca n thus to lerate numerous localized sires of congestion o r damage to capillaries or collect ing ven ules. Cessation of bloodflow and

local ischemia in the vulnerable, sub­epithelia l region will on ly occur when a ll of t he bra nches feeding a collecting venule have been damaged o r con­gested. S imilarly, the redundancy in t he

Figure 8 ) Scanning electTon microgralJh of venous plexus ( vp). This is a view of the undersurface or serosal aspecc of the mucosa. The vessels of the submucosa, muscularis and serosa were removed during /)re/Jaration, and the imal{e shows che interconnected nature of che venous plexus and the overlymg ca/Jillary bed (x50)

CAN J 0ASTROENTEROL V OL 4 No 3 M AY 1990

Figure 9) Trammission electron micrograph of a subepichelial collecting venule after exposure uf the mucosa for IO mins to 16 mM ASA in 50 mM hydrochloric acid, and for a subsequenc 5 mins to 50 mM hydrochloric acid. Nace red blood cells (rbc), parnally degranulated J1/arelet.1 ( P), and fibrin ( F) def)OS1tion (X8400)

99

MnRRIS er al

unJerlying venular plexu:, ensures that Jrainage and perfusion wi II continue even when numerous mJiviJual col­lecung venulrs have hccome 01:cluJcd. Sites of microvascular damage by topi­cally applied aspirin : When acidified aspmn b placeJ on the gastric mucma it produce:,, within minutes, extensive subepithelial platelet thrombi in the collecting venu les (Figure 9). Ag­gregareJ platelets actively synthesize thmmboxanc A2, which b a potent vasoc.l1btor anJ melf ,1 promoter of fur­ther platelet aggregation. Whi ttle et al (22) showed that when local mtravas­cular synthesis of thromhoxane A2 was :,timulateJ, exposure of the canine stomach to concentrations of aciJ and hile ,alt ( I 00 mM hydrochloric dCIJ and 5 mM sodium caurocholatc), which were normally tolemtcd, resulted in ex­tensive mucosa! bleeding.

The impairment of hemosrnt1c mechanisms, which is usually aii­soc1ated with .1spirin 111gest1on, re~ults from a systemic effect, vrn 0-acetylarinn of p l.nelet cyclo-oxygenase activity. However, in the gastric anJ upper intcs­tmal muco~ac, rhe ASA molec.ulc need only penetrate a ~ingle layer of epithelium and a fcv\ microns of ex­tracellular mamx before commg into contact with the entlot helia l cells of the subep1chel1al capillaries and rnllecung venule:,. ASA can therefore rapidly Jamage primarily the vascular endo­thelial cell and inhihtt prostacyclin syn­thc~is. Thromhus formation, congestion and local tissue c.h1mage, in the form of petechrne or punctate erosions, result. Effects of pa ren teral administration of N SAIDs and hypotens ion : Parenteral adm1n1stration of aspirin and inJo­methacin caused antral ulcers in the stomachs of animals when Joscs were used that gave plasma levels considered therapeutic in humans. The ulcers ap­peared in advance of signs of a broken barrier and were not related to the presence of salicylatcs in the gast ric lumen, although the presence of luminal acid was necessary for le ion formation (23). Furthermore, Whittle (24) demonstrated that whi le neither 2 mM soJ1um taurocholate, nor any one of four suhcuraneously administered

100

NSAIDs ( indomethacm, flurbiprofen, ASA, or na proxen), nor the vaso­constnctor noradrenaline, produced lesions, the comhination nf any nf the NSAIDs or noradrenaline w it h acidified taurocholate was highly ul­cerogen1c. NSAlDs prevented the hy­peremic response which norma lly follows topical Jamage. The result was the production of hemorrhagic eros11ms by a barrier breaker which would nor­mally have been tolermcJ.

The present authors found that when ASA ( 16 mM in 50 mM hyJrochloric acid) was placcJ in the ex vivo gastric chamber for up to 30 mms, only small punctare erosions devclopeJ (unpublisheJ data). Similarly, mduc­t ion of hemorrhagic shock by withJrawal of blooJ to produce a rapid hut transient Jecreasc in mean hlood pressure (to ahout 25 mml lg) produced few lesions and only temporary decrease in transmucosal potential difference. Howe,·er, when ASA was placed m the chamhcr for IO mins shortly aftL'r I he inJuctinn ofhcmorrhagk shock, exten­sive ulceranons covering .in a\'erage of 50% of the glanJular mucosa re,ultcd. These lcsiorn. first appeared as patchc~ ti intense foca I pallor and were covereJ with ,1 dense coagulum nf cxtruJed cell~ and mucus. This material uwcreJ regions of Jeep mucosa I necrosis wh Kh ultunmcly hecnme hemorrhagic m the presence of luminal acid. Such ulcers appear in the ,1hsence of diffuse damage and their gene 1~ cannm readily be derived from consideration of t he har­rier-breaking hypothesis. They probab­ly result from va~cular block.ide, perhaps due co vasoconstriction at the mucosa I base m m rhe submucosa. Effects of concentrated ethano l on the mucosa) microcirc ulation : The present morphological studies suggcH that ethanol-induced hemorrhagic erosions resulted from congestion or occlusion at the point where the mucosa! venular plexus is drained hy veins which pass through the mu:,cularis mucosa (Figure 10). This is a vulnerable site in the mucosa I circulation. Rlockadc hy com­pression or congestion of these draining veins woulJ affect relatively large areas of the mucosa. After exposure to con­centrntcd ethanol, the sites which were

Figure I 0) Sccmnmg dectmn 1111crogmph of lweral 1•1rt11 nf parual rrrrogratle cmr of mucosa/ venous ,1,rt'm Anows mdicare wm.s wluch Jrnm blood from rowus plexm mto large rnh muco~al wms (x2.'i)

destined to ht'u>tnc hemorrhag1l cros1om were clearly v1s1hlc because of t hci r congested hlood \'l'SsC Is. Suh­mucosal .irtencs, mucosa! cap1ll,mcs ,md coll cu mg w1m were I yp1cal ly con­gested nt these sites (9), hut vl'ins in the suhmuco,a bclm\ the exit from the mus­culans mucosa were empty of hlood.

Oates .md 1-Iakkincn (25) have sug­gested a sequence of events which may lead to the production of cchanol-111-duced gastric b,iom. T hey suggested chat eth,mol may ,let by causmg n1,1st

cells to rapidly dcgr,mulatl' anJ relense vasoactive mcd1ntors such as leukotncnc (LT) C4 and histamine. These mediators could then mducc venous crnv,t riu1on and gast nc muscle contractHms. The} ohsl·rveJ th.it ,irtenal and arteriolar J ilannn occurred within 15 10 20 s of cxpnsurc of the stomad1 to ethanol, and that the com­hmat1on of strung ,1rtcnal/ arteriolar dilation and vl'no u~ const r iction proJuced vascul,1r engorgement and visihle hyperemia w1thm l 111111. Greatly elevated capi lla ry pres~ure in the mucosa anJ cnhm1tcd rcrmcabi I it y can produce muco~al edema with a rcsul­ta nr li fting and b l istering of the epithelium. The authors' stud ies arc in general agreement with thcst· concepts (26,27). T he primary m ies for mast cells anJ the producuon of L TC4, however, remain 111 douht, smce ethanol caused significant mast cel l degranulauon m cytoprotected mucosa ( 28) and there

was no correlation hcrwecn the ahil11y of various cytoprotcctive agent~ to m-

CAN J GA~TROENTEROl VOL 4 No 3 MAY 1990

h1b1t LTC4synthe~i!>anJ 1he1 rability to r rotect the mucosa (29) .

LOCAL AND REFERRED CYTOPROTECTION:

PROST AGLANDIN-INDEPENDENT AND PROST AGLANDIN­

DEPENDENT A phenomeno n which the author~

h a ve J efined as ' refe rred cy to­prmecu on' occurs at snes in the mucosa whic h arc no t in contac t with the protective agent ( 30). In thc~e studies, an ex vivo rat gastric chamber prepa ra­t ion was used (27 ). This preparation allows direc t viewing of the mucosa and moni toring o f physmlogical indica tors of mucosa I integrity during the periods in which protective agents are present and during the development of he mor­rhagic erosinns. By t ilting the gastric cha mber preparati11n at an angle of a bout 30'\ it is possible to cover half of the muco~a l surface wid1 1.0 mL of lumina l solution. A 5 min exposure of half of the mucosa to the mild irrita nt -0.25 M hydroc hloric ac id, for instance - prod uced referred cywprotec tio n agamst 40% ethanol. There was a high ­ly sigr, rficam reduc uon in the area of he morrhagic erosions on horh the s ide whic h was exposed to the irri ta nt and the side whic h was not exposed. T h e a u thors have s tud ied the e ffects of severa l cytoprotec ti ve agents, and some , suc h as 4% sodium c hl oride, produced only local protec tion aga inst etha nn l, while o thers produced refe rred prot ection ( Willi a mso n c t al , manuscript submitt ed)

Th b model a llowed the examina­tion of cha nges which occurred in the hal f nf the mucosa not exposed to ir­ritan t, and testing of the hypo thesis that adaptive cycoprmect ion results from an increase in re lease of mucus or bicar­bo n a te sec re ti o n. C onse qu e ntl y, ne ithe r of these changes occurred nn the uncovered side wh ich, although not direc tly exposed to t h e protecci ve agents, was still protec ted against 40% ethanol. ln addi tion, there wa~ no in ­creased mucus re lease and no change in the th ickness of the Juxtamucosal un­s tirred/alka line layer.

Referred prntecLion against etha nol was, however, hlocked by pa rentera l in-

domethacin (5 mg/kg) and dopa mine D1 and D2 receptor antagonists. These studies suggest tha t re ferred protection ts prostaglandin-de pendent and that it mvolves dopamine rgic neurons. Local protection o n the side covered by the irritant nr prostaglandins was no t af­fec ted by mdome thac in or dopamine receptor antagornsts. The prostaglandin­dependence of referred rrotectio n a nd the ability of irntants w produce local protection in the presence of inh i hit ion of cyclo-oxygcnase may e xplain much o f the controversy in the literature abou t the subject of whe ther ad<1ptivc cyroprotection against ethano l depends on prostagland in synthesis. Referred protection depends on prosraglandin synthesis; local protection does not.

Referred protect ion was nm, how­ever, produced by e ither irritants o r prostagland in~ aga inst ac idified ASA or high concent ration (80 mM) sodium taurochola te, even though a ll of the milJ mi cants and prostagland ins wluch were studied prov ided local protec t ion agai nst th ese necrotizmg agents. Fur­the rmore, the irritants produced effec­tive local prntection even in a nimals pre treated wit h 5 mg/kg indnmeth<1c 111. The fa il ure to produce referred protec­t ion aga inst ASA and sodium taurocho­la tc probahly re fl ec ts the d iffe ring means by whic h the»e agents produce hemorrhag ic les ions compa red with con centra t ed et h a n o l. Eth a n o l produce!> vascula r occlusion m, a result of effects on vesse ls at the mucosa I hase or in the suhmucosa . The initiating even t may he a const rictio n of dra ining venule!> nr ve ins, as a result of e ither eth ano l-111duced mediator rclea!>e m cont ract ion of vessels of the muscularis muctha. Acidified ASA and sodi um wu rocholme pmducc damage which 1s 1n1 t iall y severe 1n t h e lum1n a l epithe li um, then progre!>ses into the la m mu propna, fo llowed hy an influx of luminal ac id and da mage to vascular c le ments such as the collec ting vcn ules.

PROST AGLANDINS AND THE MUCUS-BICARBONATE

BARRIER It is widely hclieved tha t a rrotecr ive

pH gradien t exists at the surface of the gastri c mucosa. T urnberg and Ross ( > J )

CAN J UASTRl)ENTER()( vm 4 NL) 3 MAY 1990

Prostaglandins and mucosal defences

have !>hown that the pl-! at the mucosa I surface can be as high as 6. 7, even in the face of a luminal pH as low as 2.0 ( J 0 mM hydrochloric ac id). O nce the lum i­nal pl I is decreased below pH 2.0, how­e ver, the Juxtamucosal pl [ gradie nt diminishl.!s. A t a lumina l pH of 1.1 (80 mM hydrochloric ac id) the gradient i!> comple te ly abo lished . These findings are of cons1dera hle interest and have done much to ~upport the idea tha t the pl I gradien t is essential fo r mucosa! de­fence and tha t the ahility of prosta­g landins to st imulate bica rbo n nte secretio n is a major factor in explainmg the ir cytoprotect iveabilities. l lowever, it 1s difficult to understand how a pl I gradien t could affect entry of ethanol in to the mucosa, and the authors have consis tently obtained prostagbrndin -in­duced protection aga inst luminal nccro­tizing agents when the c hamhered gastric mucosa is bathed in 50 mM hydrochloric acid, which has a pH of 1.3.

THE UNSTIRRED WATER LA YER AND MUCOSAL

PROTECTION Most studies of the um tirred layer

have been concerned wich its effects on absorption . h 1s becoming increasingly likely, however, that the presence ofnn unstirred layer at the gastric mucosa! sUifacc is of importance in explaining the res istance of the mucosa to the d e ve lopment of ac ute he mo rrhagic ero~ions following exposure to necrotiz­ing agents. S timulation hy prosraglan ­dins of alka line secretio n .iml muc us release could produce an increase in the thickness of the iuxt;imucrn,a l um tirred laye r suffic ient to produce phys ical d1~­sipat ion and ddumm of necrot izing agents. Duane et al (32 ) useJ a carbon mo nox 1d e diffus io n tee h n iq ue to dcm(lnstrarc an unst irred layer, ap­proximately 880 µm thick, on the sur­face of rat gastric mucosa. Mixing of the lum in a l solution resulted 111 a 45% dec rease 111 the th ickness of the un­st irred layer ( to an average t hickness of about 450 µm) and significantly in­creased susceptib ili ty of the mucosa to damage by bile sa lt.

The aut hors h ave show n t ha t lumina l stasis prior to adm inistration of ac idified sodium tau rocholmc or 40%

10 1

MORRIS ec al

ethanol :,ignificantly reJuceJ che Jamage proJuceJ hy these agent:, ( 12). Pretreatment of animab with mdo­methacin aholisheJ the prmectivc ef­fects of luminal stasis, huc chis protection was resloreJ hy sucralfme. These studies suggested that unpaired prostaglandin synthesis, 1c, 88% inh1b1-tion of synthesis of 6-keto-proscagLm­din F1a, mcrea:,cJ the susceptibility of the mucosa in the same manner as J1d stirring of lummal solutions.

THE pH GRADIENT AND CYTOPROT ECTION

The authors have stu<l ieJ the effects of protective mild irritants and luminal stasis on the thickness and magnitude of the Juxtamucosal pH grad1en1 using an ex vivo rat gastric chamber model anJ antimony microclectrodes. The microclectroJes had tip diameters of 40 µm and were mounteJ tin a micromanipulator which is capable of movement in three dimen~iom anJ aJ­vancement of the microelectro<le in LO µm inc rements. An inJ 1Herent electrode was also positioned m the chamber anJ changes m hydrogen ion concentrations were recorJed m mill1-volts on a Metrohm Hemau E5 l 2 pH meter/voltmeter. Thi:, apparatus 1s ac­curate to 0.05 pH units. Micro­electrodes were calibrated Jady u:,ing scanJard :,olutions of known pH. The bcgi n n i ng of the pl l grad 1ent was JefineJ as the point at which the read­mg increased by 2 m V or more above that of the bulk solution (equivalent to a change in pH of abouc 0.05). The thickness of the pl l graJient was JefineJ as the distance between the start of the graJient and the pomt at wh ich the microelectrode toucheJ the surface of the gastric mucosa. Place­ment of the electroJe anJ contact with the mucosa! surface were determined visually with a modified dissecting microscope. The change in pH across the gradient was defined as the mag­nitude of the pH gradient.

A ll of the results described m this section were obtained while bathmg the gastric mucosa with 50 mM hydrochloric acid. The gastric chamber containe<l stirred (200 rpm) 50 mM hydrochloric aci<l during the first three

102

10 mm pcnnJs. Durmg the fourth 10 mm pernxl the lummal solution was either st1rreJ or umc1rred. At t hl· enJ of the fourth pcnoJ, the lummal solu11on was remnveJ anJ replaced with 10 mL of 50 mM hyJmchloril ac1J, anJ meas­urements of pl l gradient dcpch and magnituJe were nbcaincJ wnhm 5 mins. Each experimental group con­tamed five animals. Four measurements of pH gradient depth anJ magnituJe were obtamcJ from two sites on each of the Jorsal anJ vencral surfaces of each sromach. The sites from which meas­urements were taken were locateJ mi<l­way between the greater anJ lesser curvmures, and were equ1Jistant frnm the anterior anJ pnsterim margins of the chambered mucosa. Effects of luminal stasis o n the pH gradient: Lummal stasis ign1ficantly (P<0.01) increased the thickness of the pH gradient from 410±20 µm when the solution was stirred to 820±30 µm (Fig­ure I 1 ). In the presence of surnng, the magnituJe of the gradient was only 0.30±0.05 pH units. Luminal stasb 111-

creased the magnituJe of the gradient to l .1±0.05 pH units. However, despite this mcrease, the juxmmucosal pl I was still less than 2.5 and the mucosa was protecteJ 111 this mstnnce without the pnxlucuon of an alblme zone near the surfaces of the ep1thel1al cclb. Prel 1mi­nary J1reu measurements of the thickness of the unstirrcJ layer have been carricJ out, anJ it is likely that the present mcasuremcnls nf jux­tamucosal pH gradient arc also in­d icat 1ve of the thickness of the unstirred layer. Effects of indomethacin on the pH gradien t: In earlier experiments the authors found that mhib1tion nf pros­rnglandin synthesis hy pretreatment with mJomethacm resulted m the loss of the protectton agamst taurocholate conferred by the IO min perio<l of luminal stasis ( l 2). The authors' studies using the antimony m1croclectnxles dcmonstrnteJ that indomethacin pretreatment significantly ( P<O.O 1) reJuced the th ickness of the p H gradient by more than 50% co a mean of 360±20 µm (Figure 11). Indn­methacin also reduced the magnitude of the gradient to 0.40±0.05 pH units.

These results 111d1cate thac impaired prostaglandm synthesis and l he resulc­ing mcreascJ suscepubility of mucosa to Jamage are rit least parually due to a Jecrease in thickness of the unstirred layer/pH graJient.

Effects of inhibiting bicarbonate syn­th es is o n the j ux t a mucoi.a l pH gradient and on mucosa! susceptibility to acute damage: The authors tested the effects of the carbonic anhydrasc mhib1tor, acctazolamiJe ( 100 mg/kg suhcutaneously) on the chambered gastric mucosa. Acetazolamide pre­treatment aboltsheJ the protective ef­fcc t of lumina l stasis against taurocholace anJ significantly reduced both thickness (to 270±50 µm) anJ magnitude (to 0.25±0.05 pH units) of the juxtamucosal pH gradient. These findings suggest that bicarhonate ~ecrc­tion 1s important in mucosa! defence anJ that it 1s impaireJ by NSAID . They also suggest , however, that che beneficial effects of bicarbonate secre­tion may be due to l'ffects on the un­st trred layer as wel I as any effeus resulting from elevation of pH at the mucosa! surface since, even in the presence of unimpaireJ bicarbonate secretion, the pl I at the mucosa I surfac.e is less than 2.5 (Figure l I).

TROPHIC EFFECTS OF PROST AGLANDINS

ProstaglanJms have marked trophic effcc.rs on the gastric mucosa. Ad­ministration of misoprostol to dogs for 11 weeks proJuceJ a 36% increase m stomach weight, primarily due to in­creases in length of the upper or foveolar region of the gland, anJ reflecting a significant increase in the glanJ cell production rate (3 3 ). Similar mcrcases m 1h1ckness of the gastric mucosa, particularly in the foveolar regions, were also seen in humans after aJministrnt1on of l 5(R)-I 5-methyl prostaglanJin Ez for four months ('34 ). The hyperplasia pnxluceJ by prolonged treatment with prostaglan<lins resu lted in increases in epithelial cells which secrete bicarbonate and mucus. This is the population which responds w luminal irritants and is responsible for mucosa! defensive responses against

CAN J GASTROENTEROL VOi 4 No '3 MAY 1990

Prostaglandins and mucosal defences

1000

900

c 800

"' 700 'o 0 ... 600 0,

:r: 500 a. 0 400 .c +-a. "'

300 Cl

200

100 J 0

'"1 1.00

(I)

"O 0.75 i :E

C CJ)

0 '"j ::l:

0.25

0.00 Figure 12) Transmrrnon dercrnn m1Cmgraf>h uf cam1r1lohaccedike m·Rcmr, m., (11ul1cwed hv <1rrow,) ch.">r:racd wrch rnrfacc e/111hclurl cell , 111

antral gland, of lnmwn ga~m c mtico,o ( X6085)

"' 0 0 t :,

" ~ -; :E ii a,

0 :c a.

''J 2.5

2.0

1.5

1.0

0.5

0.0 0.05 M HCI (Unstll'TOd)

0.05 M HCI (Stll'TOd)

lndomelhocln Acelazolamlde (Unstll'TOd) (Unstlmod)

Figure I I) Effect, of .1 tirrmg the l11mmal sol11uon wit/1 a glass paddle rotating w 200 rf>m, of mhcwaneous indomethacin (5 mg/kg) and s11brnwneo11s acewzolcm11Je ( I 00 mg/kg) .on the de/)th and magnitude of the 111xwm11rn.1al fl/ I gr£ul1e111 and on the /1/-l measured withm l O µm of the lwnmal 1urface of the chambered rat gasn·ic mucosa. The gasiric chamber contained 50 mM hydrochloric acid f>r1or to and durmg all 11ieas11n'll1l'!ltS wuh mrcroelectrnde.,. All values are significantly ( P<O.O l) less than the con·e;f>rmclinR mlue for chc unstim:d lummal solurwn

Figure 13) Lighc mrcrograJ)h of chrome cy/>e m1tral ulcer fro m rm go1rrrc mucosa. This ulcer was /rre~enL eighc ,h y, afte r nrngastrrr rnfrmon of 50 mg rrrnrrrobenzcne sulphonrc acid in 50% ethanol ( l'ki1cil seer run 1tarneJ wrth wluidme blue (x53 )

agent:, wch a~ bile salts and excess luminal ac id. A lthough it can he hypothesized that the change:. rc:,ulting from prolonged administration of pros­caglandins should render the stomach more rc~istant tn ::tcute damage , the authors ::t re unaware of any published results of such experiments.

PROST AGLANDINS, MUCOSAL DEFENSES, AND

DEVELOPMENT OF CHRONIC ULCER

About four co five million North Amcricam suffer from peptic ulcer dts­casc at any given time, and as much a~ 5 w 10% of the popularion may be

CAN J GASTROENT ERO L V OL 4 No 3 M AY l 990

affected <luring an ind1V1dual l1tctimc (> 5). There are a:,soc1at1ons hcrween depressed levels of rmlSlaglanJm syn­thesb and the presence of chrome ulcer ( 36 -38). The highc, t inc id ence of ga:,tnc ulcer 1s found among per~ons who ingest large quanutic, of NSAlD, , p::t rticularly the elderly ( 39,40). There

103

MORRIS et al

is little evidence for increased in ­

cidence of gastric ulcer in alcoholics even though acute exposure to high concentrations of ethanol produced su­perfic 1al gastric damage in humans (41). Ethanol ingestion has, however, been associated with an increased in­c idcnce of duodenal ulcer (42). Moreover, the authors have shown that acute exposure to ethanol produced ex­tensive, superficial Jam,igc to the duodenum and proximal jejunum ( 19). Cigarette smokmg 1s associated wtth the presence of brn h duodenal and gastric ulcers ( 42,4 3 ), and there 1s con ­vincing evidence that smoking produces a transient ( less than 12 h) decrease in prostaglandin synthesis in the gastric mucosa ( 44 ). The effects of smoking on duodenal prosraglandin synthesis arc less clear, although the deleterious effects of smnking on heal ­ing time an<l relapse rate are well estah­lished (45,46).

Several studies suggest that replace­ment or supplemcmatio n of endo­genous synthesis of prrn,taglandins with exogenous sources may be beneficial in situations where high <loses ofNSA IDs are hcmg used . The prostaglandin E1 analogue m1soprosrol has hecn shown to produce a significant rcducti(m in NSAIO-induced gastric lesions, a benefit which has not been demonstrated by Hz receptor anta­gonists (47). In addmon, the reduction by misoprostol of indomethacin-in­duced increases in intestinal mucosa! permeability in humans ( 48) raises the interesting possibility that exogenous prostaglandim may he heneficial in the treatment of NSAID-mduced small in­tcstinnl inflammation . Cytoprotection and the transition from acute to chronic ulcer: In um­sidenng the possible mechanism~ h; which prosrngland ins may affect rlw defen sive responii<.'S of the upper gastrointestinal tract, 1t 1s necessary en consider the type of damage which is under study. The maiomy of scudics aimed at g;:11n111g an understanding of the ct iology of chronic gastric ulcer have ac tually concen trated on the study of acutl' mucosa! erosions, which are readily produced 111 animal models. These are prevented by prwr ad-

104

ministration of prostaglandins. Protec­tion is provided in both humans and antmals against a variety of damaging agents such as A SA and ethanol.

The inherent assumptton 111 these studies has been that chronic exposure to the factors wh 1ch produce acute erosions ('barrier break ers', acid, hypovolcmia) will result 111 the development of a chron1t ulcer. If pros­taglandins act to prevent th 1s acute damage, it is suggested that they will abo prevent the development of chronic ulcer. In tact, 1t 1s now apparent that repeated exposure to agents or con ­ditions which produce acute errn,1ons does not necessarily result in c.hrontc ulceration. Instead, the gastnc mucosa may develop resistance or 'tolerance' to

the ulcerogenic agents. This occurs in both humans and antmals and h,ls been demonstrated for ASA as well as for ethanol ( 49-52).

Only under cxcepticmal circumstan­ces docs prolonged exposure to ulccro­ge n ic agents produce, 111 animals, pathologies which bear so me resemblance tll human chronic ulcer. There is neither a naturally occurring analogue of human peptic ulcer nor a w 1dely accepted an11na l mode I for chronic gastric ulcer, which can he used for studies of ulcer persistence and heal­ing. The closest appmx1mattnn w a chronic model is the acetic acid model 111 which the ~erosal applica­tion of concentrated acetic acid produces a long- lasting wound cm the mucosa I surface ( 5 3 ). Although 1t has no et 1opa thul og 1c: relevanc<.' t ci human gastric ulcer, the acettc acid model has proven useful 111 some st udies of the effects of prolonged ad­m1nistrnt1on of prosrngland1m and NSAIDs on mucosa! healing (54,55). Mechanisms of development of chronic ulcer: In the authors' studies nn animal mode ls o f chronic ul ­c:erointlammatory disease, the follow-1 ng hypo theses to explain the developmenL of chronic uker have hecn mvoked. • Chronic: ulcer is initiated by acute

damage, with consequent develop­ment of acute inflammation and tn­

c reased mucosa! permeability. • Chronic inflammatton depends cm

entry into the lamma propna of a luminal antigen which is not ade­quately cleared hy the mucosa! 11n­munc system.

• Conventional therapy may produce transient hcal111g but recurrence of acute injury or enhanced per­meability in the presence of the 111-c1ting anugen will result 111 relapse. Products of microbial infections

such as Campylohacwr pylori (Figurl' 12) are a potential source of luminal anti­gens in the upper gastrointesunal tract, whereas acute erosions can have m,my causes, including the ingestion of NSAIDs. In addtt1on, recent studies hrive demonstra ted that NSAIDs produced 111c reascd permeability to

macromolecules 111 the intestine and that prostaglandins have prevented this increased permeability (48). These ef­fects occurred a1 sites in the bowel where decreased levels of luminal acid, due to the antisecretory effects of pros­taglandins, would not have ha<l a role. Can cytoprotective prostaglandins prevent the development of chronic ulcer? The hypotheses outlined above predict that agents which prevent the development of acute hemorrhagic erosions sh11u ld prevent the otherwise inevtrnbk development of chronic ul ­ccromflammatory disease. The abilities of m1soprostol to prevent the develop­ment ofNSA ID gastropathy in humam and to reverse NSAID-induced chan­ges 111 mtestmal permeability arc con­sistent with these concepts.

The authors have examined the abil1ty of prosraglandms to prevent the dcvelopmt•nt of chronic ulceration of the sromach and colon in two animal models which they have developed. They found that orogastric intuhauon of fasted ( 18 h) female Sprague-Dawley rats with l ml of 40 or 501X> ( v/v) eth­anol concain111g 50 mg of the hapten trinitrobcnzenc sulphonic acid (eth­anol/TNB). produced one or more chronic type gastric ulcers (56). It was not necessary to use prior scmitization to

produce these chronic lesions. In this model the ccmccntrated ethanol pro­duced acute damage and the hapten TNB, when complexed with tis:,ue pro­teins, provided a p(X1rly cleared antigen.

Administration of either ethanc1l or

CAN J GA!-,TROENTEROL Vo, 4 No 3 MAY 1990

10

9

8

7

Ulcers 6

per 5 stomach

4

3

2

0 Saline Sucralfate

100 mg

**

Rioprostil 100 ug/Kg

Figure 14) Effecr.s of tn·etreat:mem with a prnsr:aglandin EI analogue. w1th sucralf ace and wnh saline ·vehicle, on the incidence of chronic type gastric ulcers present: m rat stomachs five da'ls after m·ogastric intuba1ion with '>O mg of trinitrobenzene ~ulphornc acid m ahmlute ethanol.** P<0.01

ethanol{TNB proJuced, within l h , ex­tensive, linear, he morrhagic erosions in the glanJular mucosae of a ll a nimab. There was no difference at ei the r I h or one Jay in the severity or appearance of the lesion~ proJuccJ hy ethano l or et hanol/TN 8. The .1cute d am,ige produced hy ethano l alone was totally healed, however, by about fi ve Jays after ad ministrat ion, whereas 80% of the animals which receive~I ethanol/ TNB haJ one or more chronic type u leers in the antrnl mucosa of the lesser curvature at least five to IO days .1fter administration (Figure l3). The ulcers were ovoid or c ircular anJ ranged in d iameter from 2 to 5 mm. Linear, fissur­ing ulcers were also present in the cor­pus of some animals. The antra l ulcers were assoc iated with e rythematous regions of gastritis, a nd open ulcers per­sisted for up to J 5 day~.

ACKNOWLEDGEMENTS: This re­search was supporred by the Medical Re­search C.iuncil of Canada.

REFERENCES I. Robert A, Nezamis JE, Phill ip, JP.

lnhibit10n of gastric secretion by rrosraglandins. Am J Dig Di, 1967; I Z: 1073-6.

Z. Miller TA. Protective effects of prosta-

In one set l)f exper im ents, the authors examined whether pretreat­ment of rats with cytoprotective agent~ would prevent the development oft he c h ronic t ype gastric ulcers.Two groups of l O rats each were fasreJ for 18 h and then given l ml of either sa line or sal ine con tc1 ining 5 mg/kg of 16,16-di­methyl prosraglanJin Ez (dmPGE2 ), fo llowed 20 mins !mer by 50 mg of TNB in I mL l 00% ethano l. After seven days the a nimals were sacrificeJ, and the number and size of ch ronic type ulcers in the gast ric mucosae were deter­mined. Pretreatment with dmPGEz re­duced the mean lesion number from 4.8 per animal to zero. In a s imilar experi­ment in wh ich the effects o f sucra lfate ( lOO mg) and the PGE1 an a logue rio­prost il ( 100 µ g/kg) were tested for the ir ability to prevent gastric ulcers, only rioprostil produced a s ignificant de-

glandins aga inst mucosa I damage: Cur­rent knowledge and proposed mechanisms. Am J Physiol 1983;245:G601-23.

3. Robert A, Nczamis JE, Phill ips JP. Effect of prostaglandin Et on gastric

secretion and ulcer formation in the rm. Gastroenterology 1968;55:481-7.

4. Rohen A, Schultz JR, Nczamis JE, Lancm,ter C. Gastric antisecrccory ant i­ulcer properties(>( PGEz, 15-methyl PGE2. and 16, 16-dimethyl PGE2.

CAN J GASTROENTEROL VOL 4 No 3 MAY l 990

Prostaglandins and mucosa! defences

crease at Jay 5 after intubanon with 2 mL 100% ethanol containing 50 mg TNB (Figure 14) . In experiments on a moJel for ch ronic, transmurnl inflam­mation of the colon, the authors found that rioprostil, when given 20 mins he­fore inrracolonic infusion with etha­no l{TNB, also s ignificantly reduced the severity of the resulting c hronic inflam­mation ( 57 ). In these models, the protec­tive effects of the prosraglandin,~ probably reflected their abi lity to prevent acute dmnage, and thus to decrease entry of the reactive hapten (TNB) in to the lamina propria (58).

CONCLUSION It is becoming appa re nt that the

gastrointestinal mucosa relies for its de­fence on a series of dynamic responses. The henefic ial e ffec ts of proscaglan­d ins cann ot be en tire ly exp lained with reference to their effec ts on only one aspect of this defensive rcper­ll)irc. The abilit ies of prostaglandins rn suppress acid secret ion and modify NSAID-induced changes in mucosa l permeability cou ld aiJ in both the healing of chronic ulcer a nd preven­t io n of relapse. Effects on mucosa! bloodflow, mucus release a nd a lkaline secretion coulJ explain resistance to

the deve lopment of acute erosions and accelera t eJ recovery fro m acute damage. Similarly, protection aga inst acute d amage can also result from trophic effects and prevention by pros­tagland ins of the high amplitude gastric contraction:, which result from exposure to various ulcerogenic procedures and agents. These 'cyLOprotective' effects may, by preventing acute injury, also prevent the initial stage necessary for the pnxluction of chronic ulcer or for relap~e of healed ulcer.

lnrravcnou,, oral and intrajcjunal ;id­minbtration. G;istroentcrology J 97 6; 70: 3 59-70.

5. Rohert A, Nczamis JE, Lancaster C, Handar J. Cytoprotectinn by prosrn­glandin in rats. Prevention of gaMric necrosi; produced by alcohol, HCI, NaOl-1, hypcrtonic NaCl, and thermal injury. Gastmentcrology 1979;77:431-4 3.

6. Rohen A, Nczamis JE, LancaS1cr C, Dav1~Jr. Field SO, Hanchar AJ. Mild

105

MnRRl<.,eral

irritants prevent gastnL necm;,1s ,uu 1t1n biopsy and hy acute ethanol m- MR. Lcvm S. Allen JL. Wright NA. through 'adapriw ..:y1oprotenion' take in normal hum,1n sm,111 intc,tine. ProsraglanJins and 1hl' ga,tm med1 ,1t,·d hy rr,,,1agland111,. A111 J Dig [)1, Sc i 1980; Vi:51 >-2 5. cpirheliu111 : Effcc1s ,if n11sopn"tol on Physinl 198 >;245:G 11 1-21. 20. G,mnon B. Rrnwnmg J. O'Brien P. gastric ep1tlwli.il ce ll prolill'ral 11m 11)

7. Guth Pl I, Paubl'n (~. Aspirin induced Rogers P. Muuisal milmvascular archi - the dog. Gut 1989;30: 316-2 I. ga.,trk injury 111 the rar: I lbtol,igK ll'C\Ure of I he fund us ,md h,Kly Of 34. Tyrgat GN, Offorhaus GJA, Van changes anJ sucralfate cyroprotect1011 . human stt 1111ach Ga,rnicnterology Minnen AJ, Everts V, I k·nscn-Log• Prue Soc Exp Biol Med 1987;1 84:421-8. I 984;86:866-75 man, SC, S::ims,m G. Influence of nrnl

8. Domschkc W, I lagcl J. Ruppm 11, z I. G.mnon I\ BrnwningJ, l)'Hnen P. l 5(R)· LS-methyl prrn,taglanJin Ez ,m Knduk R. AnwciJs and gastric mucosa! T he m1cruvasuilar an:hir,·uure nf the human gastric mucosa. A light micro· protection. Scand J Ga,tmememl glandular rnucns,1 ,if the rat ,t,H11,1ch. ,copic, cell kinetic, and ultrnstructur:il 1986;2 I (Suppl 125): 144-lJ. J Anat 1982;115:667-81. srudy. GiistrtX'ntcrology 1986;90: 111 1-20.

9. Mnrri, UP. Keenan CM, Shriver DA 22. Whittlc FIJR. K.1utfm,m GL. M<mc.1d,1 l5. Goldherg MA. Medical treatment of Morrht1log1cal and phy,1olug1.:.1l S. Vasoums1nc1nn wirh rhmmhox,111c peptic ulcer disease: ls it truly eff1c.1-ctkcc, tif a cywpmtcct1vc pro,u- Az mdu<.Cs ulcl•rnrton of rhc gast nL rnlu,,? AmJ Med 1984;77:589-91. glandm analng (noprostil) on rhc rar mucos.1. Nature 1981;292:4724. 16. H1ll1cr K, Smith CL, Jc.:we ll R, Arthur gastric muuisa. Cl111 ln\'cst Med n Kauffman (JL. Grossman Ml. Prost,\· MJP, Ross G . Duodenal mucosa syn-1987;10: 121-3 1. ghmdin and cimettdint> inh1hit the for- thes1., nf prosraglandins in , luodcnal

10. I lawh·y CJ, Kemp RT. Walt RP. 111,1tion nf ulcers produced hy ulcertlise,1\c. Gut 1985;26:237-40. Bhaskar NK, Davies J, Fil1pow1c: Fl. parcntcr,11 salicylatcs. Gastrnenterology 37. Cnm1pron JR, Gihhorn, LC, Rees EvidenLl: that .1darr1,·e cyt,>pwtect 10n 1978;7 5: 1099-102. WDW. Simulrancoll', measurement of 111 mt, 1, not mediated hy prosta· 24. Whittle RJR. The potcnrimion nf in vitro gastroduo<lcm1l prosrnglandin gland111s. Uastrnentemlogy rn11rnchohue-111duced rat gastric Ez synthesis and degradation in pcptK I 988;94:948-54. erosinns fo llowing parenteral admin- ulcer dbeasc. Scam! J G,1scruentcrol

II . MarNaughton WK, Will iamson TE, 1srrnt1nn nf cyclooxygen:ise mhib1rors. I 987;22:425-10. Morris G P. Adapt 1ve cywprotection Br J Pharmacol ! 983;80:545-51. 38. Pugh S. Will iams SE, Lewin MR, er al. by 0.25 M I !Cl 1s truly 'cytopmtect1vc' 25. Oates PJ, Hakkmen JP. Studies on the Duodenal and ancrnl mucos,11 pr,l,-and m,iy nm J epend upon cle\'atcd mechanism of eth:inol-induced gastric rnglandin Ez synthesis 111 a study of nor-lcvcb nf pro,taglandm synthcs1,. Cm J damage in rnts. Gastrocnterology mal subjects and all , rages of duodenal Phys1ol Pharmacol 1988;66: I 075-8 1. 1988;94: L0-2 1. ulcer disease treated by I '2 rccepror

12 M(irris GP, Keenan CM, 26. Morris GP, Wallace JL. The roles of antagonists. Gut 1989; 30: L 61-5. MacNaughton WK, Wallace JL, ethanol anti o( acid in the pro<lucrion 39. Faulkner G, Prichard P, Somerville K, Williamson TE. Protection of rat of gastnc mucosa! erosions in rats. Langman MJS. Aspirin and bleeding gastric mucosa by sucralfatc: Effects of Virchows Arch [BJ 1981; 38:23-38. peptic ulcers m the elderly. Br McJ J lumin,ul stasi, ,rnd l)f mhihition nf 27. Wallace JL, Morri:; GP, Krausse EJ, 1988;297: 13 11-l rrosmglandin synrhem. Am J Med Greaves SE. Reduction by cytoprotcc- 40. Langm,m MJS. Anti-inOammawry I 989;86(Suppl 6A): I 0-6. rive agents of ethanol-induced damage drug intake and the risk <if ulcer com-

13 . Chemer JA, Naik L, Tamawski A, to the rat gastric mucosa: A correlare<l plications. Med Toxicol 1986; I (Suppl Br:owwski T. Srachurn J, Smgh CJ. mnrpholog1cal and phys1olog1cal study. I ):34-8. Al:11liry of prom1glandin to reduce Can J Physiol Pharmacol 41. Gort fried EB, Korsten MA, Lieber CS. eth,rnol injury to d1,pcrscd ch1ei cells L 982;60: 1686-99. Alcohnl-induced gastric and duo<lenal from guinea pig stomach. Am J Phys1ol 28. Beck PL, Morris G P. Wallace JL. lesion, in man. Am J Gasrroencerol I 989;256:G704- l 4. Reduction of ethanol-imluced gastric I 978;70:587-92.

14. Tcr,1110 A, H1raish1 I l, Orn S, damage hy sodium cromoglycarc and 42. Piper DW, Nasiry R, McIntosh J, Shy Sug11nom T. Cytoprotccrion hy prosra • FPL-5 2694. Role of leukomcnes, CM. Pierce J, Byth K. Smoking, glandms in cell culture . In: Haya1shi prosragland111s, and mast .:ells 111 clw alcohol. am1lgesics, and chronic 0, Yamamnto S, eds. Advances in Pr,"· prorccuvc mechanism. C,m J Physiol duodenal ulcer. Scand J G;istroenrerol raglandin, Thromboxanc and Pharmawl I 989;67:287-9 >. 1984; 19: IO I 5-2 1. Lcukornenc Research. New York: 29. Wallace JL, Beck PL, Moms GP. 4 l. McIntosh JH, Byth K, Piper DW. Raven Pre.s, 1985:655-6. Is there a role for leukotnenes as Environmental factors in c1ct1ology of

15. T cmno A, Mach T, Sw .. hur.1 J, mediators uf ethanul-mduced gastric chrunic gastric ulcer: A case control Tamawski A, Ivey KJ. Effect of 16.16- mucosa! damage? Am J Physiol study of exposure variables before the Jimethy l rrustaglandin E2 on aspirin-in- 1988;254:G I 17-23. first sympwms. Gut 1985 ;26: 789-98. duced damage tll rat gastric epithclml >O. Donaldson CL. Morris GP. Local and 44. McCready DR, Clark L, Cohen MM. cell~ in tissue culrure. Gut 1984;25: 19-25. 'referred' cytoprotcction hy mild Cigarette smoking reduce, human

16. Szabo S, Hollander D. Pathw11ys of 1mrants and a POE1 an,ilog. Castro- gastric lumin:il prosraglandin Ez. gastroinresrim1l protection and repair: ent.:rology 1988;94:A614. (Absr) Gut 1985;26:1192-6. Mechanisms of act iun of sucrnlfatc. 31. Turnberg LA, Ross IN. Studies of the 45. Guslandi M. How docs smoking harm Am J Med 1989;86(Suppl 6A):2 3-31. pH gradient :icross gastric mucus. the duodenum? Br Med J

17 Morri, G P. Prosrnglandins and cellular Scand J Gastrnenrerol 1984; 19 1988;296: 111-2. restitution in the gastric mucosa. Am J (Suppl 92):48-50. 46. Hull DH, Beale PJ . Cigarette smoking Med 1986;8l(Suppl 2A):23-9. 12. Duane WC, Levitt MD, Staley NA, and duodenal ulcer. Gut 1985;26: I 131-7.

18. Ivey KJ, Baskin WN, Krnusc WJ, Terry Mel !ale AP, Wiegand SM, Fetzer CA. 47. Graham DY, Agrawal NM, Roth SR. B. Effect of aspinn and acid on human Role of the unstirred layer in protect· Prevention ofNSAID-induced ga,cric Jl'junal mucosa. An ultrn,tructurnl mg the murine gastric mucosa from ulcer with misoprostol: Mulcicentrc, study. Gastroenten>lllgy 1979; 76:50-6. bile s:ilt. Gastroentcrology double-blind, placebo-controlled trial.

19. Millan MS, Morrb GP, Beck IT, 1986;9 l :911-8. Lancet L988;i i: 1277-80. Henson JT. Villous damage induced by n. Goodlad RA, Madgwick AJ, Moffatt 48. Bjamson I, Smethurst P, Fenn CG,

106 CAN J GASTROENTEROL VOL 4 No 3 MAY 1990

Lee CE, Menzies IS, Levi AJ. Miso­prosrol reJ uces indomethacin-induced changes in human small intestina l per­meability. Dig Dis Sci l 989;34:407- l 1.

49. Graham DY, Smith JL, Spjut 11) , Torres E. Gastric u<.h1ptation. Studies in humans during continuous aspirin ;1Jministration. Gastroen terology 1988;95: 327-33.

50. Tcppcrman BL, Soper BD. Prosrnglandin E2 binding sites in 1x1rcine oxyntic mucosa: Effects of salicylates. Can J Physiol Pharmacol l 986;64:515-20.

51. St John DJB, Yeomam ND, McDermott FT, de Boer WGRM. Adaptation of the gastric mucosa to

repeated administration of asp irin in the rar. Dig Dis Sci 197 3; 18:88 l -6.

52. Ivey KJ, Tarnawski A, Stachura).

Werner H, M.ich T, Burks M. The induction of gastric mucl)sal tolerance to ,1lcohol by ch ronic administration. J Lab Clin MeJ l 980;96:922- 32.

5 3. Okahe S, Pfeiffer CJ. The acetic acid ulcer mtxlel. A procedure for chronic duodenal or gastric ulcer. In: Pfeiffer CJ, ed. Peptic Ulcer. Ph iladelphia: JB Lippincott, l 971: 13-20.

54. Okahe S, Takeuchi K, Honda K, Takagi K. Effects of aceryl~alicylic acid (ASA), ASA plus L-glutamine and L-glutamine on healing of chronic gastric ulcer in the rat. Digestion 1976; 14:85-8.

55. Wang JY, Yamasaki S, Takeuchi K, Oka be S. Delayed healing of acetic acid-induced gastric ulcers in rats by indomethacin. Gastroenterology 1989;96:393-402.

CAN J GASTROENTEROL VOL 4 No 3 MAY 1990

Prostaglandins and mucosa! defences

56. Moms GP, Rehc1ro L, Hcrrn.lgc MS, Szcwczuk M, Depew WT. An animal 111txlcl for chronic granulommous mflammarion of rhe stomach and c,il,m. Gasrroenrcrology 1984;86:A I 188. ( A hst)

57. Wal lace JL. MacNaugh tnn WK, Morris GP, Beck PL. lnh ihit i,in of leukotr ienc synthesis markedly accelerate, hea ling in a rar m,xiel of inflammarnry bowel J1scasc. Gastrocnternlogy I 989;96:29- 36.

58. A llgayer 1-1 , Dcschryver K. Stenson WF. Treatment wirh 16, I 6'-d 11nerhyl pmsraglandin Ez heforc and after inductton of colitis with trini crohen­zenc sul fonic acid in rats Jecreases inflammation. Gasrroentcrology 1989;96: 1290-300.

107

Submit your manuscripts athttp://www.hindawi.com

Stem CellsInternational

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

MEDIATORSINFLAMMATION

of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Behavioural Neurology

EndocrinologyInternational Journal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Disease Markers

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

BioMed Research International

OncologyJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Oxidative Medicine and Cellular Longevity

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

PPAR Research

The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014

Immunology ResearchHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Journal of

ObesityJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Computational and Mathematical Methods in Medicine

OphthalmologyJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Diabetes ResearchJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Research and TreatmentAIDS

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Gastroenterology Research and Practice

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Parkinson’s Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttp://www.hindawi.com