- Email: [email protected]
Effect of indomethacin on gastric acid and bicarbonate secretion in humans
GASTROENTEROLOGY
1984;87:1339-43
Effect of Indomethacin on Gastric Acid and Bicarbonate Secretion in Humans MARK FELDMAN
and THOMAS
J. COLTURI
Department of Internal Medicine, University of Texas Health Science Center at Dallas and the Research Service. Dallas Veterans Administration Medical Center, Dallas, Texas
The role of endogenous prostaglandins in the physiologic regulation of gastric secretion is unclear. We evaluated the effect of indomethacin, an inhibitor of endogenous prostaglandin synthesis, on basal gastric secretion in humans using a two-component model for calculating gastric acid and bicarbonate secretion. After a control, gastric secretory study, 11 healthy volunteers were given 50 mg of indomethacin orally every 8 h for a total of 10 doses, after which the gastric secretory experiment was repeated. Indomethacin significantly (p < 0.05) increased basal gastric juice volume, hydrogen ion concentration, osmolality, and acid output. Indomethacin increased acid secretion significantly (from 4.9 t 1.2 to 7.4 + 1.7 mmoU75 min, p < 0.02)without affecting gastric bicarbonate secretion [control 2.7 -+ 0.8, indomethacin 3.0 t 0.7 mm01175 min; p > 0.05). The increase in basal acid secretion after indomethacin administration was quite variable from subject to subject and was unaccompanied by significant changes .in basal serum gastrin concentrations. Unlike basal acid secretion, indomethacin had no significant effect on acid secretion stimulated by intragastric infusion of homogenized food. Moreover, indomethacin did not prevent intravenous somatostatin 14 from inhibiting food-stimulated acid secretion, in contrast to a previous study in rats in which indomethacin blocked the inhibitory effect of somaReceived January 17, 1984. Accepted June 30, 1984. Address requests for reprints to: Mark Feldman, M.D., Research Service (1511, Dallas Veterans Administration Medical Center, 4500 South Lancaster Road, Dallas, Texas 75216. This work was supported in part by grants AM16816 and AM17328 from the National Institutes of Health and from the Research Service of the Veterans Administration. The authors thank Dr. John H. Walsh for performing plasma gastrin measurements, Dr. Craig Brater for measuring plasma indomethacin concentrations, and Dr. Phillip Raskin for providing synthetic human somatostatin. They also acknowledge the technical assistance of Tina Barnett, Mary Walker, Julie OliverTouchstone, and Peter Chew and thank Vicky Slagle for preparing the manuscript. 0 1984 by the American Gastroenterological Association 0016-5085/84/$3.00
tostatin on acid secretion, Assuming the effect of indomethacin is due to reduced endogenous prostaglandin synthesis, we conclude that (a) in some individuals endogenous prostaglandins suppress basal acid secretion by a mechanism independent of the hormone gastrin; (b) endogenous prostaglandins play little, if any, role in the regulation of basal bicarbonate secretion by the stomach; and (c) endogenous prostaglandins do not regulate food-stimulated acid secretion, nor do they mediate the inhibitory effect of somatostatin on gastric acid secretion in humans. Many different prostaglandins reduce gastric acid secretion and stimulate gastric bicarbonate secretion when given exogenously (1-6). However, the physiologic role of endogenous prostaglandins in the regulation of gastric acid and bicarbonate secretion is unclear. Recently, we developed and validated a method for measuring gastric acid and bicarbonate secretion simultaneously in the intact stomach in vivo (6). This method is derived from a two-component model of gastric secretion and uses measurements of gastric juice volume, hydrogen ion concentration, and osmolality. We used this method to evaluate the effect of indomethacin, an inhibitor of endogenous prostaglandin synthesis, on gastric acid and bicarbonate secretion in humans. Our hypothesis was that if endogenous prostaglandins reduce gastric acid secretion or increase gastric bicarbonate secretion, or do both, then indomethacin might produce opposite effects and either increase gastric acid secretion or reduce gastric bicarbonate secretion. The effect of indomethacin on food-stimulated gastric acid secretion, measured by in vivo intragastric titration (7), was also studied. Finally, because indomethacin has been reported to prevent somatostatin from inhibiting acid secretion in the rat (8)) implying that somatostatin may inhibit acid secretion via endogenous prostaglandins, we also evaluated the effect of indomethacin on somatostatin-inhibited gastric acid secretion in humans.
1340
FELDMAN
AND COLTURI
GASTROENTEROLOGY
Methods Subjects and Consent Eleven healthy volunteers, 6 men and 5 women (average age 33 yr, range 20-51 yr), participated in these studies. Each subject was studied after a 12-h fast and each gave written informed consent for the studies which were approved by a Human Studies Research Committee. Using standard methods (9), basal acid output and peak acid output to 6 pg/kg subcutaneous pentagastrin was measured in each subject in the preliminary studies. These values averaged 2.1 ? 0.6 and 29.3 2 2.1 mmolih, respecparticipated in studies evaluating tively. All 11 subjects the effect of indomethacin on basal gastric acid and bicarbonate secretion. Six subjects participated in studies evaluating the effect of indomethacin on food-stimulated gastric acid secretion, and 6 subjects participated in studies evaluating the effect of indomethacin on somatostatininhibited gastric acid secretion. In 9 additional subjects, basal gastric acid output was measured on 2 separate test days. Mean basal acid output averaged 3.41 L 0.98 mmol/h on the first study and 3.32 t 1.19 mmolih on the second study, indicating that mean basal acid output on no medication is nearly constant despite the fact that basal acid secretion varies from day to day in individual subjects (9).
Jndomethacin Indomethacin (Indocin) was purchased from Merck Sharp and Dohme, West Point, Pa. After a control acid secretofy study was carried out on day 1, subjects ingested 50 mg of indomethacin every 8 h for 3 days. Then, after an overnight fast, subjects ingested 50 mg of indomethacin at 7 AM on the fourth day. Thus, each subject ingested a total of 500 mg of indomethacin divided into 10 eqtial doses. No subject reported any side effects from indomethacin administration. Gastric secretory studies after indomethacin administration started at 9 AM on the fourth day and plasma was obtained for determination of indomethacin concentration at this time. Plasma indomethacin concentration, measured by high-pressure liquid chromatography on coded plasma samples from these subjects, ranged from 1.5 to 3.7 pg/ml [mean 2.4 2 0.3 pgiml). Indomethacin was undetectable in plasma (co.05 @g/ml] in 4 subjects who had not received indomethacin. The indomethacin dose used in this study is that recommended for therapy of acute gouty arthritis or acute painful bursitis/tendinitis (10). Rane et al. (11) in their study also administered 50 mg of ind,omethacin orally every 8 h to healthy volunteers for a total of 10 doses. This dose of indomethacin reduced total body prostaglandin E synthesis by 95%. Total body prostaglandin E synthesis was inhibited almost completely at plasma indomethacin concentrations of 21 @g/ml (11). It therefore seems likely that, in our experiments, endogenous prostaglandin E synthesis was almost totally inhibited by this dose of indomethacin.
Basal Gastric Secretion
Acid
and
Vol. 87. No. 6
Bicarbonate
Subjects swallowed a nasogastric tube that was positioned in the gastric antrum under fluoroscopic guidance. Saliva was continually aspirated by a dental suction device. At 9 AM, plasma was obtained and coded for measurement of basal plasma gastrin concentration (12). Basal gastric secretion was then measured for five 15-min periods. From measurements of gastric juice volume, hydrogen ion concentration, and osmolality (6), as well as from measurement of plasma osmolality, the following were calculated as previously described: gastric acid secretion, gastric bicarbonate secretion, parietal volume secretion, nonparietal volume secretion. and bicarbonate concentration of nonparietal secretion (6).
Food-Stimulated
Gastric
Acid
Secretion
After the basal secretory study, a homogenized steak meal was infused directly into the stomach via a nasogastric tube and gastric acid secretion in response to the meal was measured for eight 15-min periods by in vivo intragastric titration to pH 5.5 as previously described (7). Six subjects were studied with and without indomethacin pretreatment. The effect of indomethacin on the inhibition of foodstimulated gastric acid secretion produced by synthetic human somatostatin-14 infusion was also studied. After the homogenized meal was infused into the stomach, somatostatin was infused intravenously in doses of 10, 50. and 100 pgih. Each dose was infused for three 15-min periods in a step-dose fashion. Four subjects were studied during somatostatin infusion with and without indomethacin pretreatment. In 1 of these 4 subjects and in 2 other subjects, 10 pg/h of somatostatin 14 was infused during the entire experiment (i.e., for nine 15-min periods] on 2 separate days, with and without indomethacin pretreatment. Thus, 10 ,ug/h of somatostatin 14 was studied in 6 subjects, with and without indomethacin pretreatment.
Statistics Indomethacin studies and compared by Wilcoxon’s sign rank co.05 (using two-sided tables) were ly significant. Values are expressed
control studies were test. Probability values considered statisticalas mean 2 1 SEM.
Results Effect of Indomethacin Secretion
on Basal
Gastric
As shown in Table 1, indomethacin significant increase in mean gastric juice hydrogen doubled
ion mean
concentration, net acid output
75 min [net acid output = bicarbonate secretion). Gastric
and from
led to a volume,
osmolality, and 2.2 to 4.4 mmoli
acid secretion minus acid and bicarbonate
December 1984
Table
INDOMETHACIN AND GASTRIC SECRETION
of Indomethacin on Mean (&SE) Basal Gastric Secretion in 11 Healthy Subjects
1. Effect
Control Volume output
11
81 2 15”
32.9 t
7.7
47.5 5 6.5”
210 +- 10
(mmoU75 min) Acid secretion (mmoU75 min) Bicarbonale secretion (mmoIi75 min) Parietal secretion (ml175 min) Nonparietal secretion
(ml175 min) Bicarbonate concentration nonparietal
of
on the
Indomethacin
60 t
(ml175 min)
Hydrogen llon concentration (mmol/L] Osmolality (mosmolikg) Net acid output
ministration, there were again no increases second challenge.
1341
230 2 7”
2.2 2 0.9
4.4 + 1.2”
4.9 + 1.2
7.4 + 1.7”
2.7 5 0.8
3.0 2 0.7
31 + 8
46 2 11”
30 + 5
35 ?z 5
88 * 10
83 + 12
secretion
(mmol/L)
Effect of Indomethacin Administration on Food-Stimulated Gastric Acid Secretion As shown in Figure 2, indomethacin pretreatment had no effect on food-stimulated gastric acid secretion. Total 2-h acid secretion averaged 38.8 t 5.8 mmol on the control day and 39.8 ‘- 5.9 mmol after indomethacin administration (p > 0.05). As shown in Figure 3, somatostatin infusion inhibited food-stimulated gastric acid secretion in a doserelated manner. Indomethacin pretreatment had no effect on inhibition of acid secretion produced by somatostatin. For example, 50 pgih of somatostatin inhibited food-stimulated acid secretion by 80% ? 10% (range 52%-93$,), whereas after indomethacin pretreatment this same dose of somatostatin inhibited acid secretion by 83% ? 10% (range 53%-97%).
a p < 0.05 vs. control.
Discussion secretion rates, as well as parietal and nonparietal volume secretion rates, were then calculated from gastric juice volume, hydrogen ion concentration, and osmolality, and from plasma osmolality (which averaged 292 “_ 1 mosmolikg). As indicated in Table 1, indomethacin increased gastric acid secretion significantly, but had no significant effect on gastric bicarbonate secretion. Likewise, indomethacin increased parietal volume secretion significantly without affecting nonparietal volume secretion or the bicarbonate concentration of basal nonparietal secretion. As shown in Figure 1, the effect of indomethacin on net basal acid outp?it was quite variable. In some subjects, indomethacin administration led to a large increase in acid output, whereas in others there was little or no effect. Increased basal acid output with indomethacin administration was not associated with changes in basal plasma gastrin concentrations, which averaged 69 f 6 pg/ml on the control day and 63 -+ 6 pg/ml after indomethacin administration (p > 0.05). ‘There was no correlation between plasma indomethacin concentration and basal acid secretion (r = 0.136). Five subjects were restudied a second time before and after indomethacin rechallenge. In the 2 subjects in whom basal acid secretion had increased after indomethacin administration, acid secretion again increased with indomethacin administration, but not as much (4.1 and 4.1 mmol increase the first time: 3.2 and 0.5 mmol increase the second time]. In the 3 subjects in whom basal acid secretion did not change much the first time after indomethacin ad-
Prostaglandins are not stored within cells, but immediately diffuse from their site of synthesis. One method for studying prostaglandin-mediated responses, therefore, is by inhibiting prostaglandin synthesis (e.g., with nonsteroidal antiinflammatory drugs). In our study, the administration of indomethacin increased basal acid secretion significantly, although the effect was quite variable from subject to subject. The effect of indomethacin on acid secretion was independent of basal plasma gastrin concentrations. Our study is supported by the work of Marcolongo et al. (13~4) who iound that the administration of 190-300 mg of indomethacin per day for 25 days increased basal or histamine-stimulated acid secretion, or both, in healthy volunteers. Assuming that the effect of indomethacin was due to inhibition
BASAL ACID OUTPUT (mmol/75min)
8 ’ 4 I
INDOMETHACIN
Figure 1 Basal gastric acid output in 11 healthy
subjects studied on a control day and after administration of indomethacin 50 mg p.o. every 8 h for 10 doses. The effect of indomethacin was statistically significant (p < 0.02) but quite variable from subject to subject.
1342 FELDMAN
AND
COLTURI
,
GASTROENTEROLOGY
Vol.
137, No.
6
FOOD
40 tr FOODSTIMULATED ACID “:;my;ON
INDOMETHACIN
30 SOMATOSTATIN 20 I
0
10
0i
%M*T~sTAYN-~~ (Ag/h)
0’30 TIME
Figure
2.
60 90 120 (minutes) AFTER
Figure FOOD
Mean cumulative food-stimulated gastric acid tion after intragastric infusion of homogenized food on a control day and after indomethacin in 6 subjects. The difference between the two studies was not statistically significant.
of prostaglandin synthesis (IS), it appears that endogenous prostaglandins may suppress basal gastric acid secretion in humans. The intersubject variability suggests that prostaglandins may play a more important inhibitory role in some individuals than in others. In vitro and in vivo studies in animals have found that arachidonic acid, a precursor of prostaglandins, The inhibitory inhibits parietal cell function (16,17). effect of arachidonic acid on parietal cells can be prevented by indomethacin (16,17), suggesting that parietal cells use arachidonic acid to synthesize inhibitory prostaglandins. It has been suggested by in vitro studies that inhibitory prostaglandins reduce acid secretion by preventing histamine-mediated activation of adenylate cyclase and generation of cyclic adenosine monophosphate within the parietal cell (1619).
Unlike basal acid secretion, indomethacin had no significant effect on food-stimulated gastric acid secretion in our study. Because food-stimulated acid secretion is mainly mediated by gastrin (20), our data indirectly suggest that endogenous prostaglandins have little effect on gastrin-stimulated acid secretion. This is supported by a study by Bennett et al. (21) who found that rectally administered indomethacin did not significantly affect pentagastrin-stimulated acid secretion. Inhibition of food-stimulated and gastrin-stimulated acid secretion by exogenous prostaglandins [22) is probably a pharmacologic effect of these drugs. A study in rats has suggested that the inhibitory effect of somatostatin on acid secretion is mediated, not by a direct action on the parietal cell, but indirectly via endogenous inhibitory prostaglandins (8). This concept was based on the observation that acute indomethacin pretreatment could prevent inhibition of acid secretion induced by somatostatin.
3.
DdSOEO
Effect of indomethacin pretreatment on the inhibitory effect of intravenous synthetic somatostatin 14 on foodstimulated gastric acid secretion in 4 subjects. Three doses of somatostatin were infused-lo, 50, and 100 pgih-each for a 45-min period in a step-dose manner. Compared to a control study during which 0.15 M sodium chloride (rather than somatostatin) was infused intravenously, somatostatin reduced food-stimulated acid secretion in a dose-related fashion. Indomethacin pretreatment did not alter the inhibitory effect of somatostatin on food-stimulated acid secretion.
Our studies in humans are unable to confirm this concept, because inhibition of food-stimulated gastric acid secretion by somatostatin was unaffected by pretreatment with indomethacin. The mechanism by which somatostatin reduces gastric acid secretion in humans remains to be clarified. Because exogenous prostaglandins stimulate gastric bicarbonate secretion in several species (1,2,46), it was our intent to also study the effect of indomethacin on gastric bicarbonate secretion in humans. Basal bicarbonate secretion averaged -2 mmolih in 11 healthy subjects and was not significantly affected by indomethacin, neither was calculated nonparietal volume secretion nor the concentration of bicarbonate in nonparietal secretion. The calculated bicarbonate concentration of nonparietal secretion was much greater than plasma bicarbonate in this and a previous study (6). Thus, bicarbonate is probably actively secreted by gastric mucosa (2). Our studies using indomethacin suggest that this basal. active bicarbonate secretion occurs independent of endogenous prostaglandins. This contrasts with experiments in vitro using amphibian gastric mucosa in which indomethacin inhibited active bicarbonate secretion (23). On the other hand, our findings agree closely with in vivo studies in dogs (24) and cats (25) in which acute administration of a large intravenous dose of indomethacin had no effect on gastric bicarbonate secretion. Thus, current evidence would seem to suggest that gastric mucosal damage induced in some people by indomethacin (26-28) is not due to suppression of endogenous gastric bicarbonate secretion. Whether the increase in basal acid secretion after indomethacin administration contributes to the ulcerogenic effect of this drug is uncertain.
December
1984
References 1. Robert A. Prostaglandins and the gastrointestinal tract, In: Johnson LR, ed. Physiology of the gastrointestinal tract. New York: Raven, 1981:1407-34. In: Johnson 2. Flemstrom G. Gastric secretion of bicarbonate. LR. ed. Physiology of the gastrointestinal tract. New York: Raven, 1981:603-16. 3. Mihas AA, Gibson RG, Hirschowitz BI. Inhibition of gastric secretion in the dog by 16.16-dimethyl prostaglandin EP~Am J Physiol 1976:230:351-6. 4. Garner A, Heylings JR. Stimulation of alkaline secretion in amphibian-isolated gastric mucosa by 16,16-dimethyl PGE, and PGF,,. A proposed explanation for some of the cytoprotective actions of prostaglandins. Gastroenterology 1979: 76:497-503. 5. Bolton JP, Cohen MM. Stimulation of non-parietal cell secretion in canine Heidenhain pouches by 16,16-dimethyl prostaglandin E,. Digestion 1978;17:291-9. 6. Feldman M. Gastric bicarbonate secretion in humans. Effect of pentagastrin, bethanechol, and 11,16,16-trimethyl prostaglandin E,. J Clin Invest 1983;72:295-303. 7. Fordtran JS, Walsh JH. Gastric acid secretion rate and buffer content of the stomach after eating. Results in normal subjects and in patients with duodenal ulcer. J Clin Invest 1973; 52:645-57. 8. Ligumsky M. Goto Y, Debas H, Yamada T. Prostaglandins mediate inhibition of gastric acid secretion by somatostatin in the rat. Science 1983;219:301-3. 9. Feldman M, Richardson CT. Gastric acid secretion in man. In: Johnson LR, ed. Physiology of the gastrointestinal tract. New York: Raven, 1981:693-707. 10. Physicians Desk Reference. Oradell. N.J., Medical Economics, 1983;37:1309. 11. Rane A, Oelz 0, Frolich JC, et al. Relation between plasma concentration of indomethacin and its effect on prostaglandin synthesis and platelet aggregation in man. Clin Pharmacol Ther ‘1978:23:658-68. 12. Walsh JH. Radioimmunoassay of gastrin. In: Rothfield B, ed. Nuclear medicine in vitro. Philadelphia: JB Lippincott. 1974:231-48. 13. Lunghetti R, Bianco G. Marcolongo R, Carcassi A. Effect of indomethacin on gastric function. Settim Med 1967;55:133953.
INDOMETHACIN AND GASTRIC SECRETION
14. Marcolongo
1343
R. Undesirable side effects of indomethacin. Clin Ter 1972;60:343-53. 15. Metz SA. Anti-inflammatory agents as inhibitors of prostaglandin synthesis in man. Med Clin North Am 1981;65:71357. 16. Skoglund ML, Nies AS, Gerber JG. Inhibition of acid secretion in isolated canine parietal cells by prostaglandins. J Pharmaco1 Exp Ther 1982;220:371-4. 17. Konturek SJ, Mikos E, Pawlik W. Walus K. Direct inhibition of gastric secretion and mucosal blood flow by arachidonic acid. J Physiol (Lond) 1979;286:15-28. 18. Sol1 AH, Whittle BJR. Prostacyclin analogues inhibit canine parietal cell activity and cyclic AMP formation. Prostaglandins 1981;21:353-65. 19. Levine RA, Kohen KR, Schwartzel EH, Ramsay CE. Prostaglandin E,-histamine interactions on CAMP, cGMP. and acid production in isolated fundic glands. Am ] Physiol 1982; 242:G21-6. 20. Feldman M, Walsh JH, Wong HC, Richardson CT. Role of gastrin heptadecapeptide in the acid secretory response to amino acids in man. J Clin Invest 1978:61:308-13. 21. Bennett A, Stamford IF, Unger WC. Prostaglandin E2 and gastric acid secretion in man. J Physiol 1973;229:349-60. 22. Konturek SJ, Oleksy J, Biernat J, Sito E, Kwiecien N. Effect of synthetic I5-methyl analog of PGE, on gastric acid and serum gastrin response to peptone meal, pentagastrin. and histamine in duodenal ulcer patients. Dig Dis 1976;21:291-5. 23. Garner A, Flemstrom G. Heylings JR. Effects of antiinflammatory agents and prostaglandins on acid and bicarbonate secretions in the amphibian-isolated gastric mucosa. Gastroenterology 1979;77:451-7. 24. Konturek SJ. Tasler J. Bilski J, Kania J. Prostaglandins and alkaline secretion from oxyntic, antral. and duodenal mucosa of the dog. Am J Physiol 1983;245:G539-46, 25. Smeaton LA, Hirst BH. Allen A, Garner A. Gastric and duodenal HCO,- transport in vivo: influence of prostaglandins. Am J Physiol 1983;245:G751-9. 26. Lovgren 0, Allander E. Side effects of indomethacin. Br Med J 1964;1:118. 27. Katz AM, Pearson CM, Kennedy JM. A clinical trial of indomethacin in rheumatoid arthritis. Clin Pharmacol Therap 1965;6:25-30. 28. Beirne JA, Bianchine JR, Johnson PC, et al. Gastrointestinal blood loss caused by tolmetin, aspirin and indomethacin. Clin Pharmacol Therap 1974;16:821-5.
1984;87:1339-43
Effect of Indomethacin on Gastric Acid and Bicarbonate Secretion in Humans MARK FELDMAN
and THOMAS
J. COLTURI
Department of Internal Medicine, University of Texas Health Science Center at Dallas and the Research Service. Dallas Veterans Administration Medical Center, Dallas, Texas
The role of endogenous prostaglandins in the physiologic regulation of gastric secretion is unclear. We evaluated the effect of indomethacin, an inhibitor of endogenous prostaglandin synthesis, on basal gastric secretion in humans using a two-component model for calculating gastric acid and bicarbonate secretion. After a control, gastric secretory study, 11 healthy volunteers were given 50 mg of indomethacin orally every 8 h for a total of 10 doses, after which the gastric secretory experiment was repeated. Indomethacin significantly (p < 0.05) increased basal gastric juice volume, hydrogen ion concentration, osmolality, and acid output. Indomethacin increased acid secretion significantly (from 4.9 t 1.2 to 7.4 + 1.7 mmoU75 min, p < 0.02)without affecting gastric bicarbonate secretion [control 2.7 -+ 0.8, indomethacin 3.0 t 0.7 mm01175 min; p > 0.05). The increase in basal acid secretion after indomethacin administration was quite variable from subject to subject and was unaccompanied by significant changes .in basal serum gastrin concentrations. Unlike basal acid secretion, indomethacin had no significant effect on acid secretion stimulated by intragastric infusion of homogenized food. Moreover, indomethacin did not prevent intravenous somatostatin 14 from inhibiting food-stimulated acid secretion, in contrast to a previous study in rats in which indomethacin blocked the inhibitory effect of somaReceived January 17, 1984. Accepted June 30, 1984. Address requests for reprints to: Mark Feldman, M.D., Research Service (1511, Dallas Veterans Administration Medical Center, 4500 South Lancaster Road, Dallas, Texas 75216. This work was supported in part by grants AM16816 and AM17328 from the National Institutes of Health and from the Research Service of the Veterans Administration. The authors thank Dr. John H. Walsh for performing plasma gastrin measurements, Dr. Craig Brater for measuring plasma indomethacin concentrations, and Dr. Phillip Raskin for providing synthetic human somatostatin. They also acknowledge the technical assistance of Tina Barnett, Mary Walker, Julie OliverTouchstone, and Peter Chew and thank Vicky Slagle for preparing the manuscript. 0 1984 by the American Gastroenterological Association 0016-5085/84/$3.00
tostatin on acid secretion, Assuming the effect of indomethacin is due to reduced endogenous prostaglandin synthesis, we conclude that (a) in some individuals endogenous prostaglandins suppress basal acid secretion by a mechanism independent of the hormone gastrin; (b) endogenous prostaglandins play little, if any, role in the regulation of basal bicarbonate secretion by the stomach; and (c) endogenous prostaglandins do not regulate food-stimulated acid secretion, nor do they mediate the inhibitory effect of somatostatin on gastric acid secretion in humans. Many different prostaglandins reduce gastric acid secretion and stimulate gastric bicarbonate secretion when given exogenously (1-6). However, the physiologic role of endogenous prostaglandins in the regulation of gastric acid and bicarbonate secretion is unclear. Recently, we developed and validated a method for measuring gastric acid and bicarbonate secretion simultaneously in the intact stomach in vivo (6). This method is derived from a two-component model of gastric secretion and uses measurements of gastric juice volume, hydrogen ion concentration, and osmolality. We used this method to evaluate the effect of indomethacin, an inhibitor of endogenous prostaglandin synthesis, on gastric acid and bicarbonate secretion in humans. Our hypothesis was that if endogenous prostaglandins reduce gastric acid secretion or increase gastric bicarbonate secretion, or do both, then indomethacin might produce opposite effects and either increase gastric acid secretion or reduce gastric bicarbonate secretion. The effect of indomethacin on food-stimulated gastric acid secretion, measured by in vivo intragastric titration (7), was also studied. Finally, because indomethacin has been reported to prevent somatostatin from inhibiting acid secretion in the rat (8)) implying that somatostatin may inhibit acid secretion via endogenous prostaglandins, we also evaluated the effect of indomethacin on somatostatin-inhibited gastric acid secretion in humans.
1340
FELDMAN
AND COLTURI
GASTROENTEROLOGY
Methods Subjects and Consent Eleven healthy volunteers, 6 men and 5 women (average age 33 yr, range 20-51 yr), participated in these studies. Each subject was studied after a 12-h fast and each gave written informed consent for the studies which were approved by a Human Studies Research Committee. Using standard methods (9), basal acid output and peak acid output to 6 pg/kg subcutaneous pentagastrin was measured in each subject in the preliminary studies. These values averaged 2.1 ? 0.6 and 29.3 2 2.1 mmolih, respecparticipated in studies evaluating tively. All 11 subjects the effect of indomethacin on basal gastric acid and bicarbonate secretion. Six subjects participated in studies evaluating the effect of indomethacin on food-stimulated gastric acid secretion, and 6 subjects participated in studies evaluating the effect of indomethacin on somatostatininhibited gastric acid secretion. In 9 additional subjects, basal gastric acid output was measured on 2 separate test days. Mean basal acid output averaged 3.41 L 0.98 mmol/h on the first study and 3.32 t 1.19 mmolih on the second study, indicating that mean basal acid output on no medication is nearly constant despite the fact that basal acid secretion varies from day to day in individual subjects (9).
Jndomethacin Indomethacin (Indocin) was purchased from Merck Sharp and Dohme, West Point, Pa. After a control acid secretofy study was carried out on day 1, subjects ingested 50 mg of indomethacin every 8 h for 3 days. Then, after an overnight fast, subjects ingested 50 mg of indomethacin at 7 AM on the fourth day. Thus, each subject ingested a total of 500 mg of indomethacin divided into 10 eqtial doses. No subject reported any side effects from indomethacin administration. Gastric secretory studies after indomethacin administration started at 9 AM on the fourth day and plasma was obtained for determination of indomethacin concentration at this time. Plasma indomethacin concentration, measured by high-pressure liquid chromatography on coded plasma samples from these subjects, ranged from 1.5 to 3.7 pg/ml [mean 2.4 2 0.3 pgiml). Indomethacin was undetectable in plasma (co.05 @g/ml] in 4 subjects who had not received indomethacin. The indomethacin dose used in this study is that recommended for therapy of acute gouty arthritis or acute painful bursitis/tendinitis (10). Rane et al. (11) in their study also administered 50 mg of ind,omethacin orally every 8 h to healthy volunteers for a total of 10 doses. This dose of indomethacin reduced total body prostaglandin E synthesis by 95%. Total body prostaglandin E synthesis was inhibited almost completely at plasma indomethacin concentrations of 21 @g/ml (11). It therefore seems likely that, in our experiments, endogenous prostaglandin E synthesis was almost totally inhibited by this dose of indomethacin.
Basal Gastric Secretion
Acid
and
Vol. 87. No. 6
Bicarbonate
Subjects swallowed a nasogastric tube that was positioned in the gastric antrum under fluoroscopic guidance. Saliva was continually aspirated by a dental suction device. At 9 AM, plasma was obtained and coded for measurement of basal plasma gastrin concentration (12). Basal gastric secretion was then measured for five 15-min periods. From measurements of gastric juice volume, hydrogen ion concentration, and osmolality (6), as well as from measurement of plasma osmolality, the following were calculated as previously described: gastric acid secretion, gastric bicarbonate secretion, parietal volume secretion, nonparietal volume secretion. and bicarbonate concentration of nonparietal secretion (6).
Food-Stimulated
Gastric
Acid
Secretion
After the basal secretory study, a homogenized steak meal was infused directly into the stomach via a nasogastric tube and gastric acid secretion in response to the meal was measured for eight 15-min periods by in vivo intragastric titration to pH 5.5 as previously described (7). Six subjects were studied with and without indomethacin pretreatment. The effect of indomethacin on the inhibition of foodstimulated gastric acid secretion produced by synthetic human somatostatin-14 infusion was also studied. After the homogenized meal was infused into the stomach, somatostatin was infused intravenously in doses of 10, 50. and 100 pgih. Each dose was infused for three 15-min periods in a step-dose fashion. Four subjects were studied during somatostatin infusion with and without indomethacin pretreatment. In 1 of these 4 subjects and in 2 other subjects, 10 pg/h of somatostatin 14 was infused during the entire experiment (i.e., for nine 15-min periods] on 2 separate days, with and without indomethacin pretreatment. Thus, 10 ,ug/h of somatostatin 14 was studied in 6 subjects, with and without indomethacin pretreatment.
Statistics Indomethacin studies and compared by Wilcoxon’s sign rank co.05 (using two-sided tables) were ly significant. Values are expressed
control studies were test. Probability values considered statisticalas mean 2 1 SEM.
Results Effect of Indomethacin Secretion
on Basal
Gastric
As shown in Table 1, indomethacin significant increase in mean gastric juice hydrogen doubled
ion mean
concentration, net acid output
75 min [net acid output = bicarbonate secretion). Gastric
and from
led to a volume,
osmolality, and 2.2 to 4.4 mmoli
acid secretion minus acid and bicarbonate
December 1984
Table
INDOMETHACIN AND GASTRIC SECRETION
of Indomethacin on Mean (&SE) Basal Gastric Secretion in 11 Healthy Subjects
1. Effect
Control Volume output
11
81 2 15”
32.9 t
7.7
47.5 5 6.5”
210 +- 10
(mmoU75 min) Acid secretion (mmoU75 min) Bicarbonale secretion (mmoIi75 min) Parietal secretion (ml175 min) Nonparietal secretion
(ml175 min) Bicarbonate concentration nonparietal
of
on the
Indomethacin
60 t
(ml175 min)
Hydrogen llon concentration (mmol/L] Osmolality (mosmolikg) Net acid output
ministration, there were again no increases second challenge.
1341
230 2 7”
2.2 2 0.9
4.4 + 1.2”
4.9 + 1.2
7.4 + 1.7”
2.7 5 0.8
3.0 2 0.7
31 + 8
46 2 11”
30 + 5
35 ?z 5
88 * 10
83 + 12
secretion
(mmol/L)
Effect of Indomethacin Administration on Food-Stimulated Gastric Acid Secretion As shown in Figure 2, indomethacin pretreatment had no effect on food-stimulated gastric acid secretion. Total 2-h acid secretion averaged 38.8 t 5.8 mmol on the control day and 39.8 ‘- 5.9 mmol after indomethacin administration (p > 0.05). As shown in Figure 3, somatostatin infusion inhibited food-stimulated gastric acid secretion in a doserelated manner. Indomethacin pretreatment had no effect on inhibition of acid secretion produced by somatostatin. For example, 50 pgih of somatostatin inhibited food-stimulated acid secretion by 80% ? 10% (range 52%-93$,), whereas after indomethacin pretreatment this same dose of somatostatin inhibited acid secretion by 83% ? 10% (range 53%-97%).
a p < 0.05 vs. control.
Discussion secretion rates, as well as parietal and nonparietal volume secretion rates, were then calculated from gastric juice volume, hydrogen ion concentration, and osmolality, and from plasma osmolality (which averaged 292 “_ 1 mosmolikg). As indicated in Table 1, indomethacin increased gastric acid secretion significantly, but had no significant effect on gastric bicarbonate secretion. Likewise, indomethacin increased parietal volume secretion significantly without affecting nonparietal volume secretion or the bicarbonate concentration of basal nonparietal secretion. As shown in Figure 1, the effect of indomethacin on net basal acid outp?it was quite variable. In some subjects, indomethacin administration led to a large increase in acid output, whereas in others there was little or no effect. Increased basal acid output with indomethacin administration was not associated with changes in basal plasma gastrin concentrations, which averaged 69 f 6 pg/ml on the control day and 63 -+ 6 pg/ml after indomethacin administration (p > 0.05). ‘There was no correlation between plasma indomethacin concentration and basal acid secretion (r = 0.136). Five subjects were restudied a second time before and after indomethacin rechallenge. In the 2 subjects in whom basal acid secretion had increased after indomethacin administration, acid secretion again increased with indomethacin administration, but not as much (4.1 and 4.1 mmol increase the first time: 3.2 and 0.5 mmol increase the second time]. In the 3 subjects in whom basal acid secretion did not change much the first time after indomethacin ad-
Prostaglandins are not stored within cells, but immediately diffuse from their site of synthesis. One method for studying prostaglandin-mediated responses, therefore, is by inhibiting prostaglandin synthesis (e.g., with nonsteroidal antiinflammatory drugs). In our study, the administration of indomethacin increased basal acid secretion significantly, although the effect was quite variable from subject to subject. The effect of indomethacin on acid secretion was independent of basal plasma gastrin concentrations. Our study is supported by the work of Marcolongo et al. (13~4) who iound that the administration of 190-300 mg of indomethacin per day for 25 days increased basal or histamine-stimulated acid secretion, or both, in healthy volunteers. Assuming that the effect of indomethacin was due to inhibition
BASAL ACID OUTPUT (mmol/75min)
8 ’ 4 I
INDOMETHACIN
Figure 1 Basal gastric acid output in 11 healthy
subjects studied on a control day and after administration of indomethacin 50 mg p.o. every 8 h for 10 doses. The effect of indomethacin was statistically significant (p < 0.02) but quite variable from subject to subject.
1342 FELDMAN
AND
COLTURI
,
GASTROENTEROLOGY
Vol.
137, No.
6
FOOD
40 tr FOODSTIMULATED ACID “:;my;ON
INDOMETHACIN
30 SOMATOSTATIN 20 I
0
10
0i
%M*T~sTAYN-~~ (Ag/h)
0’30 TIME
Figure
2.
60 90 120 (minutes) AFTER
Figure FOOD
Mean cumulative food-stimulated gastric acid tion after intragastric infusion of homogenized food on a control day and after indomethacin in 6 subjects. The difference between the two studies was not statistically significant.
of prostaglandin synthesis (IS), it appears that endogenous prostaglandins may suppress basal gastric acid secretion in humans. The intersubject variability suggests that prostaglandins may play a more important inhibitory role in some individuals than in others. In vitro and in vivo studies in animals have found that arachidonic acid, a precursor of prostaglandins, The inhibitory inhibits parietal cell function (16,17). effect of arachidonic acid on parietal cells can be prevented by indomethacin (16,17), suggesting that parietal cells use arachidonic acid to synthesize inhibitory prostaglandins. It has been suggested by in vitro studies that inhibitory prostaglandins reduce acid secretion by preventing histamine-mediated activation of adenylate cyclase and generation of cyclic adenosine monophosphate within the parietal cell (1619).
Unlike basal acid secretion, indomethacin had no significant effect on food-stimulated gastric acid secretion in our study. Because food-stimulated acid secretion is mainly mediated by gastrin (20), our data indirectly suggest that endogenous prostaglandins have little effect on gastrin-stimulated acid secretion. This is supported by a study by Bennett et al. (21) who found that rectally administered indomethacin did not significantly affect pentagastrin-stimulated acid secretion. Inhibition of food-stimulated and gastrin-stimulated acid secretion by exogenous prostaglandins [22) is probably a pharmacologic effect of these drugs. A study in rats has suggested that the inhibitory effect of somatostatin on acid secretion is mediated, not by a direct action on the parietal cell, but indirectly via endogenous inhibitory prostaglandins (8). This concept was based on the observation that acute indomethacin pretreatment could prevent inhibition of acid secretion induced by somatostatin.
3.
DdSOEO
Effect of indomethacin pretreatment on the inhibitory effect of intravenous synthetic somatostatin 14 on foodstimulated gastric acid secretion in 4 subjects. Three doses of somatostatin were infused-lo, 50, and 100 pgih-each for a 45-min period in a step-dose manner. Compared to a control study during which 0.15 M sodium chloride (rather than somatostatin) was infused intravenously, somatostatin reduced food-stimulated acid secretion in a dose-related fashion. Indomethacin pretreatment did not alter the inhibitory effect of somatostatin on food-stimulated acid secretion.
Our studies in humans are unable to confirm this concept, because inhibition of food-stimulated gastric acid secretion by somatostatin was unaffected by pretreatment with indomethacin. The mechanism by which somatostatin reduces gastric acid secretion in humans remains to be clarified. Because exogenous prostaglandins stimulate gastric bicarbonate secretion in several species (1,2,46), it was our intent to also study the effect of indomethacin on gastric bicarbonate secretion in humans. Basal bicarbonate secretion averaged -2 mmolih in 11 healthy subjects and was not significantly affected by indomethacin, neither was calculated nonparietal volume secretion nor the concentration of bicarbonate in nonparietal secretion. The calculated bicarbonate concentration of nonparietal secretion was much greater than plasma bicarbonate in this and a previous study (6). Thus, bicarbonate is probably actively secreted by gastric mucosa (2). Our studies using indomethacin suggest that this basal. active bicarbonate secretion occurs independent of endogenous prostaglandins. This contrasts with experiments in vitro using amphibian gastric mucosa in which indomethacin inhibited active bicarbonate secretion (23). On the other hand, our findings agree closely with in vivo studies in dogs (24) and cats (25) in which acute administration of a large intravenous dose of indomethacin had no effect on gastric bicarbonate secretion. Thus, current evidence would seem to suggest that gastric mucosal damage induced in some people by indomethacin (26-28) is not due to suppression of endogenous gastric bicarbonate secretion. Whether the increase in basal acid secretion after indomethacin administration contributes to the ulcerogenic effect of this drug is uncertain.
December
1984
References 1. Robert A. Prostaglandins and the gastrointestinal tract, In: Johnson LR, ed. Physiology of the gastrointestinal tract. New York: Raven, 1981:1407-34. In: Johnson 2. Flemstrom G. Gastric secretion of bicarbonate. LR. ed. Physiology of the gastrointestinal tract. New York: Raven, 1981:603-16. 3. Mihas AA, Gibson RG, Hirschowitz BI. Inhibition of gastric secretion in the dog by 16.16-dimethyl prostaglandin EP~Am J Physiol 1976:230:351-6. 4. Garner A, Heylings JR. Stimulation of alkaline secretion in amphibian-isolated gastric mucosa by 16,16-dimethyl PGE, and PGF,,. A proposed explanation for some of the cytoprotective actions of prostaglandins. Gastroenterology 1979: 76:497-503. 5. Bolton JP, Cohen MM. Stimulation of non-parietal cell secretion in canine Heidenhain pouches by 16,16-dimethyl prostaglandin E,. Digestion 1978;17:291-9. 6. Feldman M. Gastric bicarbonate secretion in humans. Effect of pentagastrin, bethanechol, and 11,16,16-trimethyl prostaglandin E,. J Clin Invest 1983;72:295-303. 7. Fordtran JS, Walsh JH. Gastric acid secretion rate and buffer content of the stomach after eating. Results in normal subjects and in patients with duodenal ulcer. J Clin Invest 1973; 52:645-57. 8. Ligumsky M. Goto Y, Debas H, Yamada T. Prostaglandins mediate inhibition of gastric acid secretion by somatostatin in the rat. Science 1983;219:301-3. 9. Feldman M, Richardson CT. Gastric acid secretion in man. In: Johnson LR, ed. Physiology of the gastrointestinal tract. New York: Raven, 1981:693-707. 10. Physicians Desk Reference. Oradell. N.J., Medical Economics, 1983;37:1309. 11. Rane A, Oelz 0, Frolich JC, et al. Relation between plasma concentration of indomethacin and its effect on prostaglandin synthesis and platelet aggregation in man. Clin Pharmacol Ther ‘1978:23:658-68. 12. Walsh JH. Radioimmunoassay of gastrin. In: Rothfield B, ed. Nuclear medicine in vitro. Philadelphia: JB Lippincott. 1974:231-48. 13. Lunghetti R, Bianco G. Marcolongo R, Carcassi A. Effect of indomethacin on gastric function. Settim Med 1967;55:133953.
INDOMETHACIN AND GASTRIC SECRETION
14. Marcolongo
1343
R. Undesirable side effects of indomethacin. Clin Ter 1972;60:343-53. 15. Metz SA. Anti-inflammatory agents as inhibitors of prostaglandin synthesis in man. Med Clin North Am 1981;65:71357. 16. Skoglund ML, Nies AS, Gerber JG. Inhibition of acid secretion in isolated canine parietal cells by prostaglandins. J Pharmaco1 Exp Ther 1982;220:371-4. 17. Konturek SJ, Mikos E, Pawlik W. Walus K. Direct inhibition of gastric secretion and mucosal blood flow by arachidonic acid. J Physiol (Lond) 1979;286:15-28. 18. Sol1 AH, Whittle BJR. Prostacyclin analogues inhibit canine parietal cell activity and cyclic AMP formation. Prostaglandins 1981;21:353-65. 19. Levine RA, Kohen KR, Schwartzel EH, Ramsay CE. Prostaglandin E,-histamine interactions on CAMP, cGMP. and acid production in isolated fundic glands. Am ] Physiol 1982; 242:G21-6. 20. Feldman M, Walsh JH, Wong HC, Richardson CT. Role of gastrin heptadecapeptide in the acid secretory response to amino acids in man. J Clin Invest 1978:61:308-13. 21. Bennett A, Stamford IF, Unger WC. Prostaglandin E2 and gastric acid secretion in man. J Physiol 1973;229:349-60. 22. Konturek SJ, Oleksy J, Biernat J, Sito E, Kwiecien N. Effect of synthetic I5-methyl analog of PGE, on gastric acid and serum gastrin response to peptone meal, pentagastrin. and histamine in duodenal ulcer patients. Dig Dis 1976;21:291-5. 23. Garner A, Flemstrom G. Heylings JR. Effects of antiinflammatory agents and prostaglandins on acid and bicarbonate secretions in the amphibian-isolated gastric mucosa. Gastroenterology 1979;77:451-7. 24. Konturek SJ. Tasler J. Bilski J, Kania J. Prostaglandins and alkaline secretion from oxyntic, antral. and duodenal mucosa of the dog. Am J Physiol 1983;245:G539-46, 25. Smeaton LA, Hirst BH. Allen A, Garner A. Gastric and duodenal HCO,- transport in vivo: influence of prostaglandins. Am J Physiol 1983;245:G751-9. 26. Lovgren 0, Allander E. Side effects of indomethacin. Br Med J 1964;1:118. 27. Katz AM, Pearson CM, Kennedy JM. A clinical trial of indomethacin in rheumatoid arthritis. Clin Pharmacol Therap 1965;6:25-30. 28. Beirne JA, Bianchine JR, Johnson PC, et al. Gastrointestinal blood loss caused by tolmetin, aspirin and indomethacin. Clin Pharmacol Therap 1974;16:821-5.