Rectal Administration of Nonsteroidal Antiinflammatory Drugs. Effect on Rat Gastric Ulcerogenicity and Prostaglandin E2 Synthesis

Transcription

1 GASTROENTEROLOGY 199;98: Rectal Administration of Nonsteroidal Antiinflammatory Drugs Effect on Rat Gastric Ulcerogenicity and Prostaglandin E2 Synthesis M. LIGUMSKY, M. SESTIERI, F. KARMELI, J. ZIMMERMAN, E. OKON, and D. RACHMILEWITZ Departments of Gastroenterology and Pathology, Hadassah University Hospital and Hebrew University Hadassah Medical School. Jerusalem, Israel Oral administration of nonsteroidal antiinflammatory drugs induces gastroduodenal mucosal damage in experimental animals as well as in humans. The aim of the present study was to evaluate the effect of rectally administered nonsteroidal antiinflammatory drugs on gastric mucosal damage, as well as on mucosal prostaglandin E2 synthesis. Fasting male rats were treated intrarectally with.5 ml 1 % NaHCO a solution containing several concentrations of either indomethacin, aspirin, ibuprofen, diclofenac, ketoprofen, or sulindac and concomitantly received 1 ml of 15 mm HCL intragastrically. Control rats received intrarectally the vehicle only. After 4 h, lesions in the secretory part of the stomach were scored and mucosal prostaglandin E2 synthesis was determined by the ex vivo prostaglandin generation technique. Dose-dependent mucosal damage was observed in indomethacin- and diclofenac-treated rats. Ketoprofen damage did not show dose dependency. In sulindac- and aspirin-treated rats, as well as in controls, no damage was detected. All drugs induced a significant and comparable degree of inhibition of prostaglandin E2 synthesis. There was no correlation between the severity of the mucosal damage and the inhibition of prostaglandin E2 synthesis. The ulcerogenicity of rectally administered nonsteroidal antiinflammatory drugs is therefore not directly related to the degree of inhibition of prostaglandin E2 synthesis and is probably related to the specific chemical and pharmacokinetic properties of each individual drug. Clinical and experimental data regarding gastroduodenal damage induced by nonsteroidal antiinflammatory drugs (NSAIDs) relates mostly to the oral application of the drugs (1-8). Gastroduodenal damage has been reported to be induced also by parenteral administration of NSAIDs, especially aspirin (9-13). Acute i.v. administration of aspirin in humans did not produce gross gastric damage (1), while in the cat i.v. administration of aspirin concomitant with stimulation of acid secretion produced lesions (12). However, systematic and substantial information regarding the gastric damaging effect of various NSAIDs administered by non oral routes is still lacking. Inhibition of gastric prostanoid synthesis has been suggested as one of the major mechanisms to explain the induction of damage by NSAIDs (14). In the present study, the ulcerogenic effect of rectally administered NSAIDs on gastric mucosa, as well as their effect on gastric mucosal prostaglandin E2 (PGE2) synthesis, was investigated in rats. The correlation among the dose administered, the extent of the damage, and the inhibition of PGE 2 generation is illustrated. Materials and Methods Male rats (Hebrew University strain) were fasted for 18 h in meshed cages with free access to water. Under light ether anesthesia,.5 ml of the tested drug solution was administered intrarectally via a plastic cannula introduced 3-4 cm proximal to the anus. The drugs tested were indomethacin, 5, 1, 2, and 5 mg/kg; diclofenac, 1, 2, and 5 mg/kg; ketoprofen, 5, 1, 2, and 5 mg/kg; ibuprofen, 2, 5, and 1 mg/kg; aspirin, 5, 1, and 2 mg/kg; and sulindac, 1, 2, and 3 mg/kg. Control rats were treated intrarectally with the vehicle only (1/ NaHC 3 ). Imme- Abbreviations used in this paper: NSAIDs, nonsteroidal antiinflammatory drugs; PBS, phosphate-buffered saline; PG, prostaglandin; PGE 2, prostaglandin E 199 by the American Gastroenterological Association /9/$3.

2 1246 LIGUMSKY ET AL. GASTROENTEROLOGY Vol. 98. No.5 diately after the intrarectal treatment. 1. ml of 15 mm HCI was orally administered to all rats. Four hours later. rats were killed by decapitation. The colon was removed. opened. and rinsed with.15 M NaCI. and the mucosa was carefully examined under a magnifying lens. The stomach was removed. opened along the greater curvature. and rinsed with ice-cold.15 M NaC!. the entire stomach was examined. and the lesions in the secretory part were measured with a ruler and scored according to length as follows: 1. punctate lesion. 2. <1 mm; mm; mm; 5. >4 mm. Total lesion scores were summed up for each treatment group consisting of 6 rats. Strips of gastric mucosa (13-15 mg) were processed for determination of PGE 2 synthesis by the ex vivo prostaglandin (PG) generation technique (15) as follows: the tissue was chopped with small scissors for 5-1 s in microfage plastic tubes in 1. ml phosphate-buffered saline (.1 M. ph 7.4. containing bovine serum albumin.1% and Na Azide.1 %) (PBS) and centrifuged in a fixed-speed bench centrifuge (Eppendorf. F.R.G.) at 1. rpm for 2 s. After discarding the supernatant. the tissue was resuspended in 1. ml of PBS and vortexed for 1 min at room temperature. followed immediately by addition of 1 /lg (1 /ll) of indomethacin to each tube to inhibit any further formation of PGs. The samples were then centrifuged for 1 min and the supernatant was stored at -2 C until assayed. Determination of PGE 2 was performed by a modification of the radioimmunoassay previously reported (16). In brief..1 ml of the specific monoclonal anti-pge 2 was added either to.1 ml of standards or to the tested solutions. then.1 ml of the [H 3 )PGE 2 (45-55 cpm) was added. Following overnight incubation at 4 C. the antibody bound to [H 3 )PGE 2 was separated from the free PGE 2 by adding.2 ml of the mixture of.1 % activated charcoal (Fisher Scientific Co.. Fair Lawn. N.J.) and.1 % Dextran T-7 (Pharmacia. Uppsala. Sweden) in PBS to each sample. Following vortex. the mixture was centrifuged at 3 rpm for 1 min in a refrigerated centrifuge (4 C). The supernatant was transferred into scintillation vials to which 7 ml of scintillation fluid was then added. and counted for 1 min each in a beta scintillator counter. Prostaglandin E2 (Sigma. Petach-Tikva. Israel) was prepared as a stock solution (1 mg/ml absolute ethanol) and kept at -2 C. A working dilution (1 /lg/ml) was freshly prepared in PBS for standards ranging from.15-1 ng/ml. Specific monoclonal anti-pge 2 was purchased from Interpharm (Rehovot. Israel). Scintillation fluid was prepared by mixing 1:2 (vol/vol) lumax (Lumac. Landgraff Netherlands) and toluene (Sigma). At the end of the experiment ml of blood was drawn from the aortas of the aspirin-treated rats into heparinized plastic tubes for determination of plasma salicylate levels by the method of Hanok (17). Morphological Studies Groups of 3 rats each treated with the various doses of ibuprofen. sulindac. aspirin. and indomethacin together and a control group treated with 1 % NaHC 3 were killed 4 h after drug administration. and the stomachs were fixed in phosphate-buffered formaldehyde. From the secretory part of each stomach. three mucosal strips obtained either from regions with mucosal damage or at random from stomachs with no macroscopic damage were embedded in paraffin. and routine 5-/lm sections were prepared. Tissues were coded. routinely stained with H&E. and blindly evaluated by light microscopy. Materials Indomethacin (Sigma Chemical Co. St. Louis. Mo.). diclofenac (Ciba-Geigy. Switzerland). ketoprofen (Specia. France). sulindac (Merck Sharpe & Dohme. U.S.A.). watersoluble aspirin (Aspegic lab. Egic Joullie. France). and ibuprofen (Teva. Israel) were used. For the preparation of aqueous solution. all drugs except aspirin were dissolved in 5/ NaHC 3 (stock solution) and brought to their final working solutions by further dilution up to 1 % NaHC 3 Statistical evaluation For each treatment group. PGE 2 generation was calculated as mean ± SE. and inhibition was expressed as percent of the control group values. The dose-dependent ulcerogenic effect of the drugs was assessed by correlating the total lesion scored in each treatment group (n = 6) with the drug dose. Because the lesion scores are variables on the ordinal scale. the nonparametric rank correlation procedure of Spearman with correction for tied ranks was used in correlating the total scores with drug doses (Figure 1). Results Four hours after intrarectal administration of indomethacin. diclofenac. and ketoprofen. gastric mucosal damage was evident only in the secretory part. The extent of the damage was dose dependent except for ketoprofen. which induced maximal damage when administered in the submaximal dose of 2 mg/kg (Table 1. Figure 1). Intrarectal administration of ibuprofen induced damage only at the highest dose (Table 1). No dose of aspirin or sulindac tested induced any gastric mucosal damage. Control rats did not have any mucosal lesions. Histological examination of the stomachs of rats treated with all doses of ibuprofen. sulindac. and aspirin showed no morphological damage and were indistinguishable from stomachs of untreated control rats (Figure 2). Histological examination of the stomach obtained from rats treated with indomethacin (3 mg/kg) showed multiple superficial ulcerations of the mucosa with hemorrhage and edema around them (Figure 3). Significant and comparable inhibition of PGEz synthesis was induced by all doses of indomethacin. diclofenac. ketoprofen. ibuprofen. and sulindac (Table 1). In all drugs. the smallest dose tested induced almost maximal inhibition. and higher doses did not cause further significant inhibition of PGE z synthesis. However. there was no

3 May 199 RECTAL ADMINISTRATION OF NSAIDS 1247 INDOMETHACIN ~ L 2 2 c '" '; 1.: 1 -l o I.. DICLOFENAC mg/kg KETOPROFEN IBUPROFEN '".. ~ '" 1 1 -l mg/kg Figure 1. Dose-dependent ulcerogenic effect of intrarectal administration of NSAIDs. Groups of fasted male rats (n = 6) were treated intrarectally with.5 ml solution of 3-4 doses of either indomethacin, diclofenac, ketoprofen, or ibuprofen. Four hours later, gastric lesions were scored according to length in millimeters, as described in Materials and Methods. Dose-dependent ulcerogenic effect of each drug was assessed by correlating the total lesion score in each treatment group with the drug dose, using the rank correlation procedure of Spearman, with correction for tied ranks. The Spearman Rho values for indomethacin, diclofenac, ketoprofen, and ibuprofen are.87,.683,.637, and.647, respectively (diclofenac, p =.2; all others, p <. 1 ~ Figure 2. Normal gastric mucosa-secretory part of a rat killed 4 h after intrarectal administration of.5 ml 1 % NaHCO. (H&E, original magnification x 2 8 ~ correlation between the extent of the damage and the inhibition of PGEz synthesis; indomethacin, 3 mg/kg, induced the most severe mucosal damage, and ibuprofen, 2 mg/kg, induced no damage at all, whereas both inhibited PGEz synthesis to the same extent. Similarly, no mucosal damage was induced by sulindac, despite very effective inhibition of PGEz synthesis by all doses. In the aspirin-treated rats, the degree of inhibition of mucosal PGEz synthesis was dose dependent and also correlated with the salicylate blood levels: 7.2 ± 1., 11.4 ± 1. and 2. ± 2. mg/dl following intrarectal administration of 5, 1, and 2 Table 1. Effect of Nonsteroidal Antiinflammatory Drugs on Gastric Mucosal Damage and Prostaglandin E2 Synthesis Treatment Control Indomethacin Diclofenac Ketoprofen Ibuprofen Sulindac Aspirin Dose (mglkg) N PCE 2 generation Total lesion (pglmg wet wt) / Inhibition score 255 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±2 97 Rats were treated intrarectally with.5 ml of 1 % NaHC 3 containing the various concentrations of each drug, and 1. ml of 15 mm HCl was orally administered concomitantly. After 4 h, rats were killed, gastric mucosal damage was scored, and PCE 2 synthesis was determined, as described in Materials and Methods.

4 1248 LIGUMSKY ET AL. GASTROENTEROLOGY Vol. 98, No.5 Figure 3. Superficial ulcerations in gastric mucosa-secretory part of a rat killed 4 h after intrarectal administration of indomethacin (5 mg/kg) (H&E, original magnification x 2 8 ~ mg/kg aspmn, respectively. However, at no dose tested did aspirin induce any mucosal damage. Local colonic mucosal damage at the site of administration was not observed following application of any of the drugs at any dose. Discussion Nonsteroidal antiinflammatory drugs induce gastroduodenal damage in experimental animals as well as in patients who take the drugs orally (1-8). The macroscopic damage may vary from mild erythema to severe ulceration (5,6). Mechanisms suggested to have a role in promoting mucosal damage include disruption of the "mucosal barrier" (18,19), inhibition of mucosal prostanoid synthesis (8,2), direct toxicity of cell enzymes, increased acid secretion, and slowing of the repair processes (14,21-24). Because almost complete inhibition of prostanoid synthesis by itself is not necessarily associated with mucosal damage (25,26) and gastric lesions induced by parenteral salicylate are not primarily associated with "mucosal barrier disruption" (9), the pathogenesis of gastroduodenal damage is more complicated (14). Route of administration, dosage, drug, and metabolism, including ionic formation and ph, also contribute to the magnitude of drug-induced damage. In human subjects, i.v. administration of aspirin did not cause detectable histological damage, in contrast to oral administration (1), while in experimental animals both i.v. and intragastric aspirin administration promote gastric damage (11,12). Most clinical and experimental information regarding NSAID-induced damage is derived from studies in which NSAIDs are administered orally. Only a few studies address the issue of induction of mucosal damage by parenteral administration of NSAIDs. In- trarectal administration of NSAIDs may induce local colonic irritation and damage at the site of application (27,28). It is also evident that oral NSAID administration can induce colonic damage (2). Whether rectal administration of NSAIDs may induce gastric damage is not clearly established. This study was designed to evaluate whether rectal administration of NSAIDs may induce gastric damage. Indomethacin is a potent NSAID which induces gastric damage by either oral or parenteral application. In this study it also induced considerable damage following rectal administration in comparable doses. On the other hand, intrarectal administration of sulindac, the pro-drug of indomethacin, did not provoke any gastric mucosal damage at equivalent dosage, although it inhibited PGE 2 synthesis to the same extent as indomethacin. Diclofenac induced gastric mucosal lesions in a dose-dependent fashion. Ibuprofen induced damage only at the highest dose, while aspirin did not provoke any damage, even at its highest dose (2 mg/kg). In a previous rat study, 1 mg/kg aspirin induced severe damage when administered orally, although i.p. administration at the same dose did not induce damage (26). Inhibition of PGE 2 synthesis was significant and similar in all treatment groups, except for aspirin at the dose of 5 mg/kg. Thus it is obvious that induction of gastric lesions in the rat by various NSAIDs administered rectally does not bear a direct relationship to the degree of inhibition of gastric mucosal PG synthesis. Similar observations were previously reported in humans by Redfern et al. (29) and Levine et al. (3). In a similar fashion we have recently demonstrated (31) that in human subjects, following 1 wk of indomethacin treatment (15 mg/day), gastric damage scores did not correlate with degrees of PG synthesis inhibition. This observation may not be true for all NSAIDs. The correlation between gastric damage and prostanoid inhibition following administration of other NSAIDs such as aspirin to the human has to be further evaluated. The extent of NSAID-induced damage in the rat stomach may be related more to the individual drug structure or composition associated with its specific pharmacokinetic properties in vivo than to the inhibition of PG synthesis alone. However, this observation may not apply to other species. References 1. Langman MJS, Morgan L, Worral A. Use of antiinflammatory drugs by patients admitted with small or large bowel perforation and hemorrhage. Br Med J 1985;29: Lewis JH. Gastrointestinal injury due to medicinal agents. Am J GastroenteroI1986;81: Ritchie WP Jr. Pathogenesis of acute gastric mucosal injury. Viewpoints on digestive diseases. 1983;15: Caruso I, Bianchi-Porro G. Gastroscopic evaluation of antiinflammatory agents. Br Med J 198;28:75-78.

5 May 199 RECTAL ADMINISTRATION OF NSAIDS Lanza FL, Royer GL, Nelson RS. Endoscopic evaluation of the effects of aspirin on gastric and duodenal mucosa. N Engl J Med 198;33: Lanza FL, Royer GL Jr, Nelson RS, Chen IT, Seckman CE, Rack MF. The effects of ibuprofen, indomethacin, aspirin, naproxen and placebo on the gastric mucosa of normal volunteers. Dig Dis Sci 1979;24: Somerville K, Faulkner G, Langman M. Non-steroidal antiinflammatory drugs and bleeding peptic ulcer. Lancet 1986;1: Robert A. An intestinal disease produced experimentally by a prostaglandin deficiency. Gastroenterology 1975;69:145-' Guth PH, Paulsen G. Effect of parenteral aspirin on the gastric mucosal barrier in the rat. Digestion 1979;19: Ivey KJ, Paone DB, Krause WJ. Acute effect of systemic aspirin on gastric mucosa in man. Dig Dis Sci 198;25: Hansen D, Aures D, Grossman MI. Comparison of intravenous and intragastric aspirin in production of antral gastric ulcers in rats. Proc Soc Exp BioI Med 198;164: Bugat R, Thompson MR, Aures D, Grossman MI. Gastric mucosal lesions produced by intravenousinfusion of aspirin in cats. Gastroenterology 1976;71: Whittle BJR. Mechanisms underlying gastric mucosal damage by indomethacin and bile salts and the action of prostaglandins. Br J PharmacoI1977;6: Whittle BJR. The mechanisms of gastric damage by nonsteroid antiinflammatory drugs. In: Cohen MM, ed. Biological protection with prostaglandins. Volume 2. Boca Raton, Florida: CRe Press, 1986; Whittle BJR, Higgs GA, Eakins KE, Moncada S, Vane JR. Selective inhibition of prostaglandin production in inflammatory exudates and gastric mucosa. Nature 198;284: Bauminger S, Zor U, Lindner HR. Radio-immunological assay of prostaglandin synthetase activity. Prostaglandins 1973;4 (suppl): Hanok A. Determination of salicylates in biological fluids. Clin Chern 1962;8: Davenport HW. Damage to the gastric mucosa: effects of salicylates and stimulation. Gastroenterology 1965;49: Cooke AR. The role of the mucosal barrier in drug-induced gastric ulceration and erosions. Am I Dig Dis 1976;21: Lange S, Henche V, Malagelada J-R. Drug-induced gastric injury in man is directly related to inhibition of mucosal prostaglandin synthesis (abstr). Gastroenterology 1986;9: Hawkey q, Rampton OS. Prostaglandins and the gastrointestinal mucosa: are they important in its function, disease or treatment? Gastroenterology 1985;89: Reeves n, Stables R. Effects of indomethacin, piroxicam and selected prostanoids on gastric acid secretion by the rat isolated gastric mucosa. Br J Pharmacol 1985;86: Levine RA, Schwartzel EH. Effect of indomethacin on basal and histamine stimulated human gastric acid secretion. Gut 1984;25: Aehringhaus U, Weiler H, Peskar BA, Peskar BM. Molecular mechanisms of the gastric toxicity of antirheumatic drugs. Arch ToxicoI1984;(SuppI7): Robert A, Tabata K, Joffe SN, lacobson ED. A decrease in prostaglandin synthesis, by itself, is not the cause of mepirizoleinduced duodenal ulcers in rats (abstr). Gastroenterology 1986; 9: Ligumsky M, Golanska EM, Hansen DG, Kauffman GL Jr. Aspirin can inhibit gastric mucosal cyclo-oxygenase without causing lesion in the rat. Gastroenterology 1983;84: Levy N, Gaspar E. Rectal bleeding and indomethacin suppositories. Lancet 1975;1: Berry H, Swinson D, lones I, Hamilton EBD. Indomethacin and naproxen suppositories in the treatment of rheumatoid arthritis. Ann Rheumatol Dis 1978;37: Redfern IS, Lee E, Feldman M. Effect of indomethacin on gastric mucosal prostaglandins in humans. Gastroenterology 1987;92: Levine RA, Petokas S, Nandi I, Enthoven D. Effect of nonsteroidal antiinflammatory drugs on gastrointestinal injury and prostanoid generation in healthy volunteers. Dig Dis Sci 1988;33: Goldin E, Stalnikowicz R, Wengrower D, Eliakim R, Fich A, Ligumsky M, Karmeli F, Rachmilewitz D. No correlation between indomethacin-induced gastroduodenal damage and inhibition of gastric prostanoid synthesis. Aliment Pharmacol Ther 1988;2: Received December 23, Accepted November 9, Address requests for reprints to: D. Rachmilewitz, M.D., Department of Gastroenterology, Hadassah University Hospital, P.O. Box 12, Jerusalem 9112, Israel.