Topical Aspirin Plus HCI Gastric Lesions in the Rat

Transcription

1 Topical Aspirin Plus HCI Gastric Lesions in the Rat Cytoprotective Effect of Prostaglandin, Cimetidine, and Probanthine P. H. GUTH, M.D., D. AURES, Ph.D., and G. PAULSEN Medical and Research Service, VA Wadsorth Hospital Center, Los Angeles, California, and University of California at Los Angeles School of Medicine, Los Angeles, California The effect of representative agents of three classes of antisecretory compounds: prostaglandins, histamine H 2 -receptor antagonist, and anticholinergic agents, on acute gastric mucosa11esions produced by topical aspirin (2 mg/kg) plus HC1 (15 mm) in the pylorus-ligated rat as studied. Acid as given exogenously so as to negate any antisecretory effect of the drugs studied. Both nonantisecretory and antisecretory doses of each agent as determined by preliminary secretory studies ere employed. The prostaglandin analogue 16,16-dimethy1 prostaglandin E2, the H 2 -receptor antagonist cimetidine, and the anticholinergic agent probanthine, in both doses studied, all significantly reduced lesion formation. The H,-receptor antagonist mepyramine neither protected by itself nor enhanced the protective effect of cimetidine. Pepsin release into the gastric content increased ith increasing mucosal damage. Hoever, addition of pepsin to the gastric instillate had no effect on severity of lesions in any group, hich indicates that the increased pepsin as the result of, and not the cause of, the mucosal damage. The findings indicate that all three classes of antisecretory agents studied are also cytoprotective, i.e., they can protect against gastric mucosal injury by topical as- Received June 26, Accepted September 1, Address requests for reprints to: Dr. Paul H. Guth, VA Wadsorth Hospital Center, Gastroenterology Section (691/111C), Wilshire and Satelle Boulevards, Los Angeles, California 973. This ork as supported by NIAMDD Grant to CURE (Center for Ulcer Research and Education) and V. A Medical Research Funds. Computing assistance as obtamed from the Health Sciences Computing Facility, UCLA, supported by National Institutes of Health Special Resources Grant RR-3. The authors are indebted to Dr. Janet Elashoff for help ith the statistical analysis. The authors are grateful for the secretarial assistance of Mr. Alan L. Gutcheon by the American Gastroenterological Association pirin plus HC1 by some mechanism other than inhibition of acid and pepsin secretion. A protective action of prostaglandin on the gastric mucosa other than by inhibition of acid secretion has been reported by Robert and his colleagues.1.2 They demonstrated that several prostaglandins, in doses that do not affect acid secretion, protected against gastric damage by nonsteroidal antiinflammatory compounds' and by chemical and physical agents such as absolute alcohol,.75 N HCl,.2 N NaOH, 3% NaCl, and boiling ater.2 Robert referred to this effect as the cytoprotective property of prostaglandin as contrasted ith its antisecretory property. Similarly, Whittle 3 reported that gastric mucosal damage produced by indomethacin plus gastric perfusion ith acid and taurocholate as inhibited by the methyl analogues of prostaglandin E2. Carmichael et a1,4 5 confirmed this effect of prostaglandin and reported that the histamine H2-receptor antagonist, cimetidine, did not have a cytoprotective effect. 5 Hoever, Rees et a1,6 found that although either an H,- or an H 2 -antagonist alone had no effect, in combination they protected against taurocholateinduced increase in ionic permeability of the gastric mucosal barrier in dogs. This latter finding raised the question as to hether the combination of antagonists might also have a cytoprotective effect like prostaglandin. The aim of the present study as to determine hether, like prostaglandin, other antisecretory agents such as histamine H2-receptor antagonists (alone or ith an H,-receptor antagonist) and anticholinergic agents had cytoprotective actions. Lesions produced by the intragastric instillation of the aspirin plus HCl in the pylorus-ligated rat as the model used for this purpose. Acid as given exogenously so as to negate any protective antisecretory effect of the drugs studied.

2 January 1979 CYTOPROTECTIVE EFFECT OF PROSTAGLANDIN, CIMETIDINE, AND PROBANTHINE H +}LEq C 15 N T R 1 L (15) 5 mg/kg ip CIME TIDINE Figure 1. Effect of 16,16-dimethyl prostaglandin E2 (PGE2) and cimetidine on gastric acid secretion in the pylorus-ligated rat. In this and subsequent figures, the vertical bars represent the mean value and the horizontal line + 1 SEM for each group; the numbers in parentheses represent the number of rats in each group; and the asterisk and double asterisks represent P <.5 and.1 (t-test comparing group mean ith control group mean). Methods Sprague-Daley rats eighing 15-2 g ere used throughout this investigation. The animals ere fasted but had free access to ater for 48 hr before study. The rats ere housed in cages ith ide mesh ire bottoms to prevent coprophagy. Secretory Study Gastric secretory studies ere performed to determine doses of each drug that inhibited and did not inhibit acid secretion under the conditons that ere to be used in the subsequent experiment. Each animal as lightly anesthetized ith ether, the abdomen opened, and the pylorus ligated; care as taken not to interfere ith the blood supply to the stomach. The abdomen as then closed, and the drug to be studied as administered. Onehalf hour after recovering from anesthesia, 2 ml of physiologic saline ere instilled into the stomach by orogastric intubation. Exactly 1 hr later the rat as killed by ether overdose, the abdomen opened, the stomach removed, and the gastric content collected for analysis. The volume of gastric content as noted and hydrogen ion concentration Table 1. Effect of Pro ban thine on Gastric Secretion No. of rats H+ Pepsin JLeq mg NaCI ±24 1 ±.13 Probanthine 5 mg/kg 4 21 ±9 b.4 ±.6 b ± 12b.2 ±.7 b.2 " 6 56 ±17 b.4 ±.4 b.4 " 6 57 ±14b.6 ±O.l b.8 " 5 16 ±32.9 ±.21 o Results are given as the mean ± SEM. b P <.1 (t-test comparing group mean ith control group mean). determined by titration to ph 7. ith.2 N NaOH using an automatic titrator (Radiometer, Copenhagen). In the study ith the anticholinergic agent, probanthine, pepsin concentration also as determined ith a modified hemoglobin substrate technique. 5 The drugs and dosages studied ere: physiologic saline subcutaneously (control); 4, 2, 1, and 1 J.!g/kg of 16,16-dimethyl prostaglandin E2 (PGE 2 ) subcutaneously; 5, 1, and 2 mg/kg of cimetidine intra peritoneally; and 5, 1,.2,.4, and.8 mg/kg of pro ban thine subcutaneously. Aspirin Study As in the secretory study, under light anesthesia, the abdomen as opened, the pylorus ligated, and the abdomen then closed, and the drug to be studied as administered. One-half hour after recovering from anesthesia, 2 mg/kg of aspirin, suspended in 1% methylcellulose in 15 mm HC1, as instilled into the stomach by orogastric intubation. Exactly 1 hr later, the rat as killed by ether overdose, the abdomen opened, and the stomach removed and opened along the greater curvature. The stomach as examined for the presence of mucosal lesions by a person ho did not have knoledge of hich drug had been administered. Lesions involved only the glandular portion of the stomach, primarily the corpus, but occasionally the antrum also. The lesions ere scored as follos: petechial lesions = 1, erosions <1 mm = 2, erosions beteen 1 and 2 mm = 3, erosions beteen 2 and 4 mm = 4, and erosions greater than 4 mm = 5. The partial scores ere then summed to obtain the final total lesion score for that animal. Statistical Analysis For each sub study (corresponding to figures or table) an analysis of variance comparing all groups resulted in significance at the 5% level. The Student t-test as then used to determine hether or not a particular mean drug result as significantly different from the corresponding mean control result.

3 9 GUTHETAL. GASTROENTEROLOGY Vol. 76, No.1 7 CONTROL MEP\ se 6 5 CIMETIDINE -IP (CIM) IIJ cr 4 o u (/) z 3 o (/) IIJ..J 2 21 (9) * (17) (7) (5) (5) 1 o Noel 2.5"'9 5"'9 DOSE I kv Figure 2. Effect of the prostaglandin analogue PGE 2 and the histamine H 2 - and HI-receptor antagonists cimetidine (CIM) and mepyramine (MEP) on gastric mucosal lesions produced by the intragastric instillation of aspirin plus HCI in the pylorus-ligated rat. 7.5mQ 5"'9 Results Secretory Study Results of the secretory study ith the prostaglandin analogue and cimetidine are presented in Figure 1. The control animals secreted 242 ± 26 (mean ±SE) meq H+ in the 1.5-hr study period. The lo dose of PGE 2, 1 ltg/kg, had no effect on acid secretion, although 2 ltg/kg produced a statistically significant 44% decrease in acid secretion. The largest dose, 4 ltg/kg, produced even greater inhibition (81%). This dose, hich is the PGE 2 concentration in the stock solution, is dissolved in 1% ethanol and proved to be very irritating on subcutaneous injection. Th~refore, the 2 ltg/kg dose as used for further studies. To milligrams per kilogram cimetidine i.p. had no significant effect on H+ secretion, hereas the 1 and 5 mg/kg doses did and caused 47% and 87% reduction in acid secretion, respectively. Results of the secretory studies ith probanthine are presented in Table 1. Five milligrams per kilogram probanthine markedly suppressed both H+ and pepsin secretion. Serial 1:5 dilutions continued to demonstrate antisecretory properties until a dose of.8 rng/kg as reached. This dose had no significant effect on either H+ or pepsin secretion. For the study of cytoprotection, to doses of each agent ere selected, one in the nonantisecretory range and one in the antisecretory range: 1 and 2,ug/kg s.c. of PGE 2, 2.5 and 5 mg/kg i.p. of cimetidine, and.8 and 5 rng/kg s.c. of probanthine. Aspirin Study The effect of PGE 2 and cimetidine in both nonantisecretory and antisecretory doses, on acute gastric mucosal lesions due to topical aspirin plus HCI is shon in Figure 2. Control animals had a lesion score of 57 ± 6. Both doses of PGE 2 significantly reduced lesion formation; the lesion score as 33 ± 7 ith the nonantisecretory and 24 ± 4 ith the antisecretory dose. Similarly, both doses of cimetidine significantly inhibited lesion formation; the lesion scores ere 34 ± 4 and 23 ± 7, respectively. The H I - receptor antagonist, mepyramine, alone had no effect on lesion formation; the lesion score as 61 ± 6. Mepyrarnine plus cimetidine did significantly reduce lesions; the score as 19 ± 6. Hoever, this effect as no greater than cimetidine given alone in that dose. Results of the study ith probanthine are shon in Figure 3. Again, both doses significantly reduced lesion formation, from 61 ± 11 to 33 ± 6 and 7 ± 2, respectively. The total amount of pepsin in the gastric content of each animal also as determined. This appeared to be related to lesion formation. hich as 8.1 ±.8 rng in the control group and 5.6 :I:.7 and 2.3 ±.2 in the lo and high dose probanthine groups. There as a linear correlation beteen lesion formation and gastric pepsin content (Figure 4), r being.7. This raised the question of hether the increased gastric pepsin content as the cause of or the result of the mucosal lesions. Therefore, an additional coded, randomized topical

4 January 1979 CYTOPROTECTIVE EFFECT OF PROSTAGLANDIN, CIMETIDINE, AND PROBANTHINE CONTROL 5 a:: PROBANTHINE u en 4 z iii *..J R (8) (8) (8) NaGI ooemo 5mg DOSE/kg Figure 3. Effect of the anticholinergic agent pro ban thine on gastric mucosal lesions produced by the intragastric instillation of aspirin plus HCl in the pylorus-ligated rat. aspirin study as performed on groups of six rats ith each using the high dose of each agent. Each agent as studied tice: once ith aspirin suspended in 15 mm Hel (and 1% methycellulose) and again ith aspirin suspended in 15 mm Hel (and 1 % methylcellulose) plus 1 mg pepsin. Results are presented in Figure 5. Again, all three agents markedly and significantly reduced lesion formation. The addition of pepsin to the gastric instillate did not have any consistent effect on lesion formation in the control group or in any of the drug groups. Discussion Acid is needed for the production of aspirininduced gastric mucosal lesions/ Inhibition of gastric acid secretion by anticholinergic agents 7 or H2-receptor antagonists B g ill protect against such aspirin-induced lesions. In the present investigation, nonantisecretory doses of three classes of antisecretory agents: prostaglandins, histamine H2-receptor antagonists, and anticholinergic agents, ere studied. The doses ere chosen by first performing secretory studies under the identical conditions of the aspirin study except that physiologic saline instead of aspirin as instilled intragastrically. Although the chosen lo dose of both cimetidine and probanthine did not significantly inhibit acid secretion, the mean acid secretion ith each as slightly loer than in the corresponding control group (Figure 1 and Table 1). The question might be raised that there as a slight antisecretory effect hich as not detected by the methods employed. Hoever, in the aspirin phase of the study, the Hel needed for lesion formation as given exogenously. In spite of the presence of topical aspirin plus Hel, all three agents decreased lesion formation. This finding indicates that all three agents can protect against lesion formation by a mechanism other than inhibition of acid secretion. In Robert's terminology,' 2 they are cytoprotective. The importance of their antisecretory effect in protecting aganist lesion formation should not be denied; hoever the present study as designed to investigate only cytoprotection. The antisecretory doses of each agent ere also studied in the eventuality, hich did not occur, that the nonantisecretory dose should prove ineffective. The possibility remains that a slight, but not statistically significant, inhibition of endogenous acid secretion might have an important effect in protecting against lesion formation in the acid-secreting portion of the stomach. Endogenous acid in the depths of the glands might have an action that exz iii II :!i 5 Q I 'I~~I~I~I~~~~~ o LESION SCORE Figure 4. Correlation beteen release of pepsin into the gastric content and lesion formation produced by intragastric aspirin + HCl in the probanthine study, r =.7.

5 92 GUTHETAL. GASTROENTEROLOGY Vol. 76, No a: u en z Q en...j o NO PEPSIN ADDED ~ 1 mg PEPSIN ADDED 1 NaG I PGEz 2 )Jg/kg CIMETIDINE 5mg/kg PROBANTHINE 5mg/kg Figure 5. Effect of the addition of pepsin to the gastric instill ate (aspirin + HGl) on lesion formation. Unlike other figures, results of t tests comparing lesion scores ith and ithout pepsin in the NaGI group and in each drug group are presented (no significant differences). Lesions scores in all three drug groups, ith and ithout pepsin, ere significantly less than in the corresponding NaGI group (P <.1). Each vertical bar represents the mean result and the horizontal line + 1 SEM of six rats. ogenous acid cannot perform. Using a parenteral aspirin plus topical acid rat model, Kauffman and Grossman 'O found that the H 2 -receptor antagonist cimetidine, as ell as PGE 2, protected against antral ulceration. With a continuous 24-hr HCl perfusion rat model, Robert et a1." reported that the anticholinergic agent methscopolamine bromide protected against duodenal ulceration. These studies on antral and duodenal lesions sho that inhibition of endogenous acid secretion cannot be the only mechanism by hich these agents act. In the probanthine substudy of the present investigation, the correlation beteen pepsin output and lesion scores raised the question of cytoprotection by inhibition of pepsin release. Hoever, the final study ith added pepsin indicated that the increased pepsin output as the result of and not the cause of the mucosal damage. Johnson '2 has shon that hen the gastric mucosal barrier is broken and acid back diffuses, there is an outpouring of pepsin from the gastric mucosa. The authors' data are consistent ith that interpretation. As as reported by Kauffman and Grossman,lO inhibition of pepsin output as not the mechanism of cytoprotection by prostaglandin and cimetidine in their model either. Robert et a1." did not study the effect of exogenous pepsin on the protective effect of methscopolamine bromide in their HCl duodenal ulcer model. The cytoprotective action of prostaglandin has been previously described.' - 5.'o In the study by Kauffman and Grossman,'o a similar action for the H 2 -antagonist cimetidine as also found. In a study of cold-restraint stress lesions in the rat, Dai et a1.13 observed that the H 2 -antagonist metiamide, in an antisecretory dose, did not further inhibit the already decreased acid secretion in the stressed rat but did significantly reduce lesion formation. That finding suggests a protective effect of metiamide other than by inhibition of acid secretion. On the other hand, Carmichael et a1.,5 using a rat aspirin plus HCl model, reported a significant reduction in lesion score (using the Wilcoxon Sum of Ranks test) by prostaglandin but not by cimetidine. Hoever, further analysis of their data reveals that reduction in lesion score by cimetidine did approach statistical significance. The lesion score ithout cimetidine as 27.4 ± 2.4 and ith cimetidine 19.7 ± 3.2. There ere 2 rats in each group. When the authors compared these to data sets by a to-sided Student t test, the probability level as.6, just shy of the.5 level usually required for statistical significance. Finally, the report by Robert et a1. of inhibition of perfused HCl induced duodenal ulceration by methscopolamine bromide" suggests a cytoprotective effect of anticholinergic agents. In sum, these studies plus the present one provide evidence for a cytoprotective effect of prostaglandins, H 2 -receptor antagonists and anticholinergic agents.

6 January 1979 CYTOPROTECTIVE EFFECT OF PROSTAGLANDIN, CIMETIDINE, AND PROBANTHINE 93 References 1. Robert A: Antisecretory, antiulcer, cytoprotective and diarrheogenic properties of prostaglandins. In Advances in Prostaglandin and Thromboxane Research, Vol. 2. Edited by B. Samuels son, R. Paoletti. Ne York, Raven Press, 1976, p Robert A, Nezamis JE, Lancaster C, et al: The cytoprotective property of prostaglandins. Gastroenterology 72:1121, Whittle BJR: Relationship beteen the prevention of rat gastric erosions and the inhibition of acid secretion by prostaglandins. Eur J Pharmacol 4: , Carmichael HA, Nelson L, Russel RI, et al: The effect of the synthetic prostaglandin analog 15(R)15 methyl-pge 2 methylester on gastric mucosal hemorrhage induced in rats by taurocholic acid and hydrochloric acid. Am J Dig Dis 22: , Carmichael HA, Nelson LM, Russell RI: Cimetidine and prostaglandins: evidence for different modes of action on the gastric mucosa. Gastroenterology 74: , Rees WDW, Rhodes J, Wheeler MH, et al: The role of histamine receptors in the pathophysiology of gastric mucosal damage. Gastroenterology 72:67-71, Brodie DA, Chase BJ: Role of gastric acid in aspirin-induced irritation in the rat. Gastroenterology 63:64-61, Okabe S, Takeuchi K, Urushidani T, et al: Effect of cimetidine, a histamine H 2 -receptor antagonist, on various experimental gastric and duodenal ulcers. Am J Dig Dis 22: , MacKercher PA, Ivey KJ, Baskin WN, et al: Protective effect of cimetidine on aspirin-induced gastric mucosal damage. Ann Intern Med 87: , Kauffman GL, Grossman MI: Prostaglandin and cimetidine inhibit antral ulcers produced by parenteral salicylates. Gastroenterology 74:149, Robert A, Nezamis JE, Stoe DF: Production of duodenal ulcers by administration of hydrochloric acid. Gastroenterology 6:79, Johnson LR: Pepsin stimulated by topical hydrochloric and acetic acids. Gastroenterology 62:33-38, Dai S, Ogle CW, Lo CH: The effects of metiamide on gastric secretion and stress ulceration in rats. Eur J Pharmacol 33: , 1975