Digestive Diseases and Sciences, Vol. 35, No. 1 (January 1990), pp. 20-26
Effect of Rioprostil, a Methylprostaglandin E 1 Analog, on Basal and Stimulated Plasma Pancreatic Hormone Levels in Man O. SEGERS, MD, and G. SOMERS, MD, PhD
The effect of rioprostil, a methylprostaglandin E 1 analog on circulating pancreatic hormones was evaluated in 13 healthy male subjects. Rioprostil administration, 300 tzg twice daily resulted in a significant decrease of fasting insulin, C-peptide, glucagon, and pancreatic polypeptide. No change in fasting plasma glucose or somatostatin levels was observed. An oral glucose tolerance test induced similar increments in plasma glucose concentration before and during treatment, but a delayed rise o f insulin and C-peptide levels occurred during the administration of the drug. On rioprostil, the glucose load no longer inhibited peripheral glucagon or somatostatin. Treatment with rioprostil remained without effect on mixed meal-induced changes in plasma glucose levels and concomitant increases in insulin, pancreatic polypeptide, and somatostatin levels. I t is concluded that in healthy individuals rioprostil influences the basal and glucose-induced levels Of glucagon, insulin, and somatostatin. In healthy men this effect did not, however, result in glucose intolerance. KEY WORDS: Prostaglandins; rioprostil; pancreatic hormones; glucose tolerance.
Synthetic prostaglandin (PG) analogs recently have been introduced in the treatment of peptic gastric and duodenal ulcers for their antisecretory (1-3) and cytoprotective (4-6) properties. During the past decade, however, several reports dealt with the effect of PG on the hormone secretion of the main pancreatic islet cells, namely the A, B, D, and PP cells, thereby influencing glucose homeostasis and possibly regulating the ingestion of nutrients. In man, PGE1 and PGE2 were described to cause inhibition of glucose-induced insulin secretion (7, 8). Likewise, all nonsteroidal antiinflammatory drugs except indomethacin, which are well-known Manuscript received January 30, 1989; revised manuscript received August 31, 1989; accepted September 15, 1989. From the Department of Internal Medecine, University Hospital AZ-VUB, Brussels, Belgium. Address for reprint requests: Dr. G. Somers, Department of Internal Medecine, Akademisch Ziekenhuis VUB, Laarbeeklaan 101, B1090 Brussels, Belgium.
20
PG synthesis inhibitors, increased basal and glucose-induced insulin levels in normal subjects (7, 9-14). Plasma glucagon levels were enhanced by PGE 1 infusion (15-17). This enhancement appeared to be an indirect phenomenon mediated either by endogenously released cathecholamines or by the interaction b e t w e e n the PG and the alphaadrenoreceptor. PG infusion has been described to stimulate gastric and pancreatic somatostatin release in vitro and in vivo (18, 19). Indomethacin apparently attenuated the postprandial somatostatin response in dogs (20). Only a few reports are available on the effects of prostaglandin analogs on pancreatic B-cell function in man. A variety of PGE~ and E2 analogs have been used in these studies. Enprostil, a PGE2 analog, caused a reduced rise in insulin after a standard meal without change in plasma glucose (21). Following administration of another PGE2 analog, eiDigestive Diseases and Sciences, Vol. 35, No. 1 (January 1990)
0163-2116/90/0100-0020506.00/09 1990PlenumPublishingCorporation
RIOPROSTIL AND PANCREATIC HORMONES ther orally, intraduodenally, or intravenously, serum insulin excursions in response to glucose were decreased (22). Misoprostol, a PGE1 analog; induced a significant e n h a n c e m e n t of plasma glucose levels and reduction of insulin output after an oral glucose load but did not change the insulin secretion after an intravenous glucose load (23). Reports on the effect of the PG analogs on the other pancreatic h o r m o n e s are scarce. Solely P G E 2 analogs have been described to reduce the p e a k p l a s m a levels of pancreatic polypeptide (PP) in r e s p o n s e to a meal (24). Postprandial p l a s m a somatostatin levels in man were not altered by either exogenous P G E 1 administration or depletion by indomethacin (25). The aim of our study was to analyze the acute and m o r e prolonged effects of oral administration of rioprostil, a PGE~ analog, on the fasting and postprandial p l a s m a glucose and on the concomitant changes in p l a s m a insulin, PP, somatostatin-like immunoreactivity (SLI), and glucagon levels in healthy individuals.
MATERIALS AND METHODS Study Protocol. The study protocol had the approval of the Ethical Committee of the University Hospital Brussels. In an open study 13 healthy male volunteers received 300 Ixg rioprostil orally, twice daily during one week. The mean age of the healthy subjects was 28 + 1 years (mean -+ SE, range 23-38 years). Body weight was within 5% of normal (26). The subjects did not take any drugs except for the test medication during the study period. A glucose load of 75 g was given in basal conditions, and after 24 hr and seven days of treatment with the methylprostaglandin E 1 analog rioprostil. A mixed meal test of 550 kcal (40% carbohydrate, 30% fat. and 30% protein) was given before and on day 6 of treatment. Meal intake time was kept below 7 min. Blood was drawn for biochemical analysis after a rest period of 15 min following the insertion of the intravenous sampling needle and was collected in chilled glass tubes containing EDTA and aprotinin (Trasylol, Bayer. Leverkfisen, FRG over a 150-min period. Blood was immediately centrifuged at 4~ and plasma was stored at - 2 0 ~ until assay. Materials. Rioprostil tablets (300 ~g) were prepared by Cilag, Schaffhausen, Switzerland. During the study, possible side or adverse effects were asked for and recorded daily. On the test day rioprostil was given 1 hr before oral glucose or mixed meal. Biochemical Analysis. Commercially available kits were used for the radioimmunological determinations of plasma insulin (Pharmacia. Uppsala. Sweden), C-peptide, and pancreatic glucagon (IRE, Fleurus. Belgium). PP and SLI were measured as previously described (27, 28). All samples of the same individual were determined in one Digestive Diseases and Sciences, Vol. 35, No. 1 (January 1990)
single radioimmunoassay. Plasma glucose was measured by glucose oxidase method. Statistical Methods. Results are expressed as mean SE. Significances of differences were estimated using an analysis of variance test (ANOVA). The significance level was set at 5%. Fasting plasma glucose and hormone levels as expressed in Figure 1 are the means of the basal values scored on the days of oral glucose tolerance test (OGTT) and/or mixed meal.
RESULTS Fasting Plasma Glucose and Hormone Levels. Figure 1 shows the fasting p l a s m a glucose and horm o n e levels before any drug intake and after one day and one w e e k of rioprostil treatment. The fasting p l a s m a glucose and S L I levels did not significantly change during the trial. A minor but significant decrease in fasting insulin, and C-peptide was o b s e r v e d during rioprostil administration. A more p r o n o u n c e d decline in fasting peripheral glucagon levels was found in the rioprostil period. M o r e o v e r , fasting glucagon levels w e r e significantly lower after 24 hr than after one w e e k of treatment with the drug. In contrast, fasting p l a s m a PP levels were only significantly d e c r e a s e d after one week of administration of the PG analog. Response to Oral Glucose. Plasma glucose excursions after the oral glucose load were not affected by a 24 hr t r e a t m e n t with rioprostil (Figure 2, upper panel). H o w e v e r , the induced rise in peripheral insulin levels a b o v e basal values was significantly lower after one day of prostaglandin treatment in c o m p a r i s o n to the increase in insulin concentrations before administration of the drug (Figure 2, lower panel). The integrated increment in p l a s m a insulin from the 30th to the 60th min after oral glucose averaged 1266 _+ 219 mU/liter on rioprostil compared to 1593 + 324 mU/liter before drug intake (P < 0.05). Again, after s e v e n days o f rioprostil, no difference in p l a s m a glucose increments could be o b s e r v e d in c o m p a r i s o n with the results obtained before t r e a t m e n t (Figure 3, u p p e r panel). H o w e v e r , in one of 13 patients a late and s y m p t o m a t i c hypoglycemia was noted after 150 min. After one week of rioprostil intake a significantly delayed rise of plasma insulin and C-peptide levels occurred (Figure 3, middle and lower panels) to levels that were slightly but significantly e n h a n c e d at 60 min after the ingestion of the c a r b o h y d r a t e . Peripheral glucagon levels w e r e inhibited b y the glucose load, After one w e e k of rioprostil h o w e v e r , the already lowered basal peripheral glucagon concentrations were
21
SEGERS AND SOMERS
8-
4 E E
3
LU
09 O O :D ._1 (3
2 1
o ._i 0
E
6"
z "i
4
co z
2
10 84
0
0
z O (,9
20 84
10-
5
B
D1
D6 7
01 .... =I DI
DI
B
D6 7
2~ ~
o
1
/// ///
I/i
iii
/~/ // /
o_
E ~
))j
--
DI
D6 7
120 -
~ 1J.I
,oo
~
80
n
60"
y/
0 B
B~
D6-7
~,/,'. B
. DI
40 D6-7
B
DI
D6-7
Fig 1. Fasting plasma glucose and pancreatic hormone levels before (B) and after 24 hr (D1) and six to seven days of rioprostil treatment (D6-7).
no longer significantly influenced by the glucose load (Figure 4, upper panel). In these healthy individuals, ingestion of 75 g of glucose resulted in an erratic change in peripheral SLI within 60 min after the glucose load and in a subsequent progressive decline of the circulating hormone levels, which OGTT
10'
0.05).
E u) o o D .J ~9
5
0
0
30
6'0
9'0
120
150
0
3'0
6'0
9'0
120
150
80 60 E i
(n z
40
20 0
TIME(min)
Fig 2. Peripheral plasma glucose (upper panel) and insulin levels (lower panel) during OGTT before (open symbols) and after 24 hr of rioprostil treatment (closed symbols).
22
became significantly lower than basal values from the 60th min after glucose intake. The SLI levels indeed progressively decreased to 62 + 7% of the basal value at 150 min after glucose intake. After seven days of rioprostil treatment, no significant reduction in peripheral SLI occurred (Figure 4, middle panel), the levels of SLI being significantly higher in comparison to the levels measured before drug treatment at 150 min and still averaging 94 + 7% of the corresponding basal concentration (P < Oral glucose did not induce any significant change in peripheral PP levels prior to or during rioprostil treatment (Figure 4. lower panel). Nevertheless in the one patient manifesting hypoglycemm after 150 min. a brisk increase in peripheral PP concentration was observed. Response to Mixed Meal. The responses of plasma glucose and the different hormones studied to ingestion o f a 550-kcal meal are shown in Figures 5 and 6. Six days of rioprostil treatment did not significantly influence the postprandial increments in plasma glucose, insulin, and C-peptide. somatostatin, or pancreatic polypeptide. Before and during rioprostil treatment, plasma glucose increased slightly to a comparable peak after 30 min (Figure 5, upper panel). Plasma insulin and C-peptide concentrations displayed a similar rise (Figure 5 middle and lower panel). Incidentally, 90 min after the ingestion of Digestive Diseases and Sciences Vol. 35. No. l (January 1990)
RIOPROSTIL AND PANCREATIC HORMONES OGTT
OGTT
~ E ~-
10'
100'
8
80' Z
60'
O
4-
0
40'
._1 (5
2"
o ..J o
20
6
O
O3
0 0
3'0
8'o
9'0
1~o ~5o
100
m ..,I
O3 Z
60 40
0
_--E
15
~,~
10
3'o
6'o
9'o
~o
1~o
E . El~,O . ~
9'0
12o 1~o
o
3'o
6'0
9'o
12o 1~o
0
30
150
8
~..
6-
r3
6'0
200
12
LU a.
3'0
20' 0
=~
o
20 84
80 E
0
100
r
4-
50
2"
0
0
o
3'o
6'o
9'o
1~o 1;o
TIME(mIn)
Fig 3. Peripheral plasma glucose (upper panel), insulin (middle panel), and C-peptide levels (lower panel) during OGTT before (open symbols) and after seven days of rioprostil administration (closed symbols).
the meal the plasma glucose and concomitant insulin and C-peptide levels were significantly higher on rioprostil. In basal conditions, the meal test induced a small but significant rise of plasma SLI over the fasting level from 30 to 120 min after the start of food ingestion (Figure 6, upper panel). During rioprostil administration a significant increase in SLI above fasting levels occurred only at 30 and 120 min after the start of the meal. However, the integrated release of SLI from 0 to 150 min was not significantly affected by the treatment with the drug (Figure 6, upper panel). The test meal also induced a biphasic rise in plasma PP levels. Again this response was not significantly altered by the treatment with the PG analog (Figure 6 lower panel). Glucagon was not measured during the meal test. Adverse Effects of Rioprostil. No adverse effects with 600 Ixg rioprostil daily were observed. Two Digestive Diseases and Sciences, Vol. 35, No. 1 (January 1990)
f
f
60 9'0 TIME(min)
120
150
Fig 4. Peripheral glucagon (upper panel), SLI (middle panel), and PP levels (lower panel) during OGTT before (open symbols) and after seven days of rioprostil treatment (closed symbols).
individuals mentioned a minor change in bowel habits with more loose stools. As already mentioned, in one subject a late hypoglycemia occurred after oral glucose load. DISCUSSION
The results of the present study suggest that rioprostil administration, 600 txg daily, to healthy volunteers influences the fasting and/or the OGTTinduced peripheral levels of insulin, glucagon, SLI, and PP. Although no effect of the drug on fasting glycemia and glucose tolerance was observed, basal plasma insulin and C-peptide levels were significantly lower during rioprostil treatment. Twentyfour hours after starting rioprostil administration, insulin secretion was decreased after an oral glucose load, which is consistent with already published data on the acute, direct effect of PG and its
23
SEGERS AND SOMERS MEAL TEST MEAL TEST
25 E
E
0,.
o0 o(D
20
...1
15
(5
3'0
e'o
9'0
z'o
8'0
9'o
,2o ,5o
3'0
6'0
9'0
120
120 400
60 300
50
E m ..J
40
,",
30"
13. 100
20"
z
10
0
0 0
3'0
6o
9o
LLI Q,.
O
Fig 6. Mixed meal-induced peripheral SLI (upper panel) and PP levels (lower panel) before (open symbols) and after six days of Rioprostil treatment (closed symbols).
4 2"
0
3'0
6'0
90
120
150
TIME(min)
Fig 5. Mixed meal-induced peripheral plasma glucose (upper panel), insulin (middle panel), and C-peptide levels (lower panel) before (open symbols) and after six days of rioprostil administration (closed symbols).
analogs on insulin secretion in man (21-23). It has to be stated that the dose of rioprostil used in this study was that recommended by the manufacturer for the treatment of peptic ulcer disease in man. No data on dose-response relationships of this compound on carbohydrate metabolism and pancreatic or intestinal hormone secretion are actually available. Nevertheless, in a recent study, administration of 300 Ixg of rioprostil, 25 min prior to an OGTT, was reported to remain without effect on basal insulin levels, but apparently caused a delayed increase in insulin levels without any changes in glucose tolerance (29). Part of the decline in insulin response observed after the administration of prostaglandin analogs was thought to be attributable to an inhibitory effect of the drugs on the
24
150
TIME(min)
8
I,Q,.
200
intestinal release of gastric inhibitory polypeptide (GIP), thus by abolishing its incretin effect (21, 22, 29). At the dose of 300 Ixg rioprostil indeed inhibited GIP secretion following an OGTT (29). Whether rioprostil administration affects other additional incretin factors such as glucagon-like peptide remains unknown (30). However, in light of the inhibitory action of prostaglandins demonstrated in vitro on B-cell function, a direct effect of the PG analog could well be at the origin of the changes in insulin output. This study further demonstrates that some effect of rioprostil on insulin excursions still persists after one week of treatment. During an OGTT, insulin and C-peptide secretion indeed displayed a slower increase, but the amplitude of the insulin response appeared to be unaffected. This delayed secretory pattern was probably the basis of the late hypoglycemic symptoms reported by one subject. The fasting pancreatic glucagon secretion was also significantly lowered during rioprostil treatment. This seems at variance with some reported data (15, 16). The reasons for the different findings could be related to the different types and routes of administration of the PG used in the different studies or to a specific action of this particular PG analog. Indeed different effects of PGE 1 and PGE2 on hormone secretions have been demonstrated Digestive Diseases and Sciences, Vol. 35, No. 1 (January 1990)
RIOPROSTIL AND PANCREATIC HORMONES
(29). Whether the observed decrease in peripheral glucagon levels results from a direct effect of the prostaglandin analog on the A-cell secretion cannot be answered by this study. However, the concomitant decline of plasma glucagon could balance the inhibited insulin output during rioprostil treatment, thereby contributing to the lack of any apparent disturbances in plasma glucose concentrations. Moreover, peripheral glucagon following an oral glucose load remained at the level of this decreased fasting glucagon concentration. The basal and stimulated plasma SLI did not change during the study. Similar findings were reported during misoprostol treatment, another PGE~ analog (25). Oral glucose has been reported to either stimulate or inhibit peripheral SLI levels (31-33). In our experience, no significant rise in circulating SLI occurred after a glucose load in healthy individuals. On the contrary, a significant decline was observed. This inhibitory effect of glucose on peripheral SLI levels appeared to be blunted during rioprostil treatment. Indeed, higher SLI levels were observed from 120 min on after the glucose intake, concomitant with the decline of the stimulated insulin levels. The interpretation of this data is difficult and may not point to any specific alterations in pancreatic D-cell function. First, only a minor, although significant, difference induced by rioprostil could be detected. Second, somatostatin is not only released into the circulation by the pancreatic D cells, but also by somatostatinproducing cells of the gastrointestinal tract and is also located in intestinal nerve fibers (34, 35). One could speculate, however, that this lack of decline in SLI somehow could play a role in the maintenance of normal glucose excursions in the presence of altered peripheral and presumably portal insulin and glucagon concentrations. The fasting PP secretion was decreased only after prolonged PG treatment. The mixed meal test after six days of treatment could not demonstrate any difference in secretory reaction of the pancreatic islet PP cell, This is in contrast with the findings with a PGE2 analog in duodenal ulcer patients in whom an inhibitory effect on stimulated PP cell secretion was observed (24). It can be advanced that part of eventual metabolic effects of prostaglandin analogs could be due to the influence of these drugs on gastric emptying, intestinal mortility, glucose absorption, or even on gastrin release (36-38). Alterations of gastric emptying apparently primarily influence postprandial Digestive Diseases and Sciences, Vol. 35, No. 1 (January 1990)
glucose levels (39). A more rapid presentation of carbohydrates at the level of the duodenum indeed can result in higher blood glucose concentrations (39). In our study, no such effect could be observed. Furthermore, no influence of rioprostil on basal and stimulated gastrin levels could be demonstrated for the rioprostil dose of 600 ~g daily (3). Likewise, rioprostil administration had no effect on the mealinduced secretion of PP and somatostatin, pancreatic hormones that apparently play a role in the regulation of either food intake, absorption or digestion (40-42). Nevertheless, it has to be mentioned that at the dose used in our study a small but significant increase in gastric emptying has been reported (43). Oral administration of rioprostil for one week was well tolerated by the healthy volunteers and no serious side effects on bowel habits were noted at the dose used. In conclusion, rioprostil treatment lowers insulin, C-peptide, PP, and glucagon levels; these alterations did not, however, induce glucose intolerance in healthy volunteers. The delayed rise of peripheral insulin levels during a glucose load could, however, result in more pronounced changes in glucose metabolism in either non-insulin-dependent diabetic patients or in individuals with glucose intolerance and an already altered insulin output. These findings suggest that careful attention should be given to such patients treated with this PG analog. ACKNOWLEDGMENTS
The authors thank Gaby Schoonjans for his technical assistance and Nadine Van Slycke for helping with the manuscript. REFERENCES 1. Akdamar K, Agrawal NM, Ertran A: Inhibition of nocturnal gastric secretion in normal human volunteers by misoprostol: A synthetic prostaglandin El, methyl ester analog. Am J Gastroentoerol 77:902-904, 1982 2. Wilson DE, Quadros E, Rajapaksa T, Adams A, Noar M: Effects of misoprostol on gastric acid and mucus secretion in man. Dig Dis Sci 31(suppl):1265-1295, 1986 3. Schulte K, Singer MV, Eysselein V, Demol P, Goebell H: Effect of rioprostil, a synthetic prostaglandin E 1, on meal stimulated gastric acid secretion and plasma gastrin levels in humans. Digestion 36:162-167, 1987 4. Robert A: Cytoprotection by prostaglandins. Gastroenterology 77:764-767, 1979 5. Detweiler MK, Harrison CA, Rollins DE, Tolman KG, MacCormick GM, Simon DM: Effects of riopostil on aspirin induced gastrointestinal mucosal changes in normal volunteers. Gastroenterology 86: P1062, 1984 (abstracts)
25
SEGERS AND SOMERS 6. Agrawal NM, Godiwala T, Arimura A, Dajani EZ: Cytoprotection by a synthetic prostaglandin against ethanol-induced gastric mucosal damage. A double blind endoscopic study in human subjects. Gastrointest Endocrinol 32:67-70, 1986 7. Robertson RP, Chen M: A role for prostaglandin E in defective insulin secretion and carbohydrate intolerance in diabetes mellitus. J Clin Invest 60:747-753, 1977 8. Giugliano D, Torella R: Prostaglandin E1 inhibits glucoseinduced insulin secretion in man. Prostaglandin Med 1:165166, 1978 9. Field JB, Boyle C, Remer A: Effects of salicylate infusion on plasma insulitis and glucose tolerance in healthy persons and mild diabetics. Lancet 1:1191-1194, 1978 10. Micossi R, Pontiroli AE, Baron SH, Tamayo RC, Lengel F, Bevilacqua M, Raggi U, Norbiato G, Foa PR: Aspirin stimulates insulin and glucagon secretion and increases glucose tolerance in normal and diabetic subjects. Diabetes 27:1196-1204, 1978 11. Chen M, Roberston RP: Effects of prostaglandin synthesis inhibitors on human insulin secretion and carbohydrate tolerance. Prostaglandins 18:557-567, 1979 12. Syv/ilahti EKG: The effect ofindomethacin on serum growth hormone, immunoreactive insulin, and blood glucose levels of young adult males. Int J Clin Pharmacol 10:111-116, 1974 13. Widstr6m A: Influence of indomethacin on glucose-induced insulin response in normal man-role of prostaglandins in the rapid insulin release? Horm Metab Res 9:172-175, 1977 14. Topoi E, Brodows RG: Effects of indomethacin on acute insulin release in man. Diabetes 29:379-382, 1980 15. Giugliano D, Torella R, Sgambato S, D'Onofrio F: Effect of ct and [3 adrenergic inhibition and somatostatin on plasma glucose, free fatty acid, insulin, glucagon, growth hormone responses to prostaglandin E 2 in man. J Clin Endocrinol Metab 48:302-308, 1979 16. Giugliano D, Torella R, D'Onofrio F: Prostaglandins and the alpha-cell. Prostaglandins Med 6:283-297, 1981 17. Sacca L, Perez G: Influence of prostaglandins on plasma glucagon levels in the rat. Metabolism 25:127-130, 1976 18. Safouri B, Weir GC, Bitas KN, Makhlouf GM: Gastrin and somatostatin secretions by perfused rat stromach: Functional linkage of antral peptides. Am J Physiol 238:495-501, 1980 19. Schusdziarra V, Rouiller D, Jaffe BM, Harris V, Unger RH: Effect of exogenous and endogenous prostaglandin E upon gastric endocrine function in dogs. Endocrinology 106:16201627, 1980 20. Schusdziarra V, Rouiller D, Harris V, Wasada T, Unger RH: Effect of prostaglandin E2 upon release of pancreatic somatostatin-like immunoreactivity. Life Sci 28:2099-2102, 1981 21. Nicholl CG, Carolan G, Sevelius H, Bloom SR: Enprostil reduces the post-prandial rises in insulin and glucosedependent insulinotropic peptide and delays small intestinal transit time. Diabetologia 29:575A, 1986 22. Konturek SJ, Mikos Em, Krol R, Wierzbicki Z, Dobrzanska M: Effect of methylated prostaglandin E2 analogue on insulin secretion in man. Prostaglandins 15:591-602, 1978 23. Ebert R, Hielscher A, Creutzfeldt W: Disruption of the enteroinsular axis by misoprostol. Eur J Clin Invest 17:A3, 1987 24. Petersen W, Feldman M, Taylor I, Brenner M: The effect of 15(R)-15-methyl prostaglandin E 2 on meal-stimulated gastric acid secretion, serum gastrin, and pancreatic polypeptide in duodenal ulcer patients. Dig Dis Sci 24:381-384, 1979
26
25. Lucey MR, Clark ML, Fairclough PD, Wass JAM, Webb J, Davidson K, Rees LM, Dawson AM: The effect of a prostaglandin agonist and antagonist on postprandial plasma somatostatin and gastrin in man. Gastroenterology 86:1167, 1984 26. The Metropolitan Life Insurance Company. Statistical Bulletin 45, 1959 27. Pipeleers DG, Pipeleers-Marichal MA: A method for the purification of single A, B, and D cells and for the isolation of coupled cells from isolated islets. Diabetologia 20:654663, 1981 28. Somers G, Pipeleers-Marichal M, Gepts W, Pipeleers D: A case of duodenal somatostatinoma: diagnostic usefulness of Calcium-pentagastrin test. Gastroenterology 85:1192-1198, 1983 29. Nauck M, Harter S, Ebert R, Creutzfeldt W: Effects of four orally administered analogues of prostaglandin E1 and E2 on glucose tolerance and on the secretion of pancreatic and gastrointestinal hormones in man. Eur J Clin Invest 19:298305, 1989 30. Kreymann B, Ghatei MA, Williams G, Bloom SR: Glucagon-like peptide 1 7-36: A physiological incretin in man. Lancet 2:1300-1304, 1987 31. Pimstone B, Berelowitz M, Kranold D, Shapiro B, Kronhelm S: Somatostatin-like Immunoreactivity in human and rat serum. Metabolism 27(Suppll):1145-1149, 1978 32. Conlon JM, McCulloch AJ, Alberti KGMM: Circulating somatostatin concentrations in healthy and non-insulin-dependent (type II) diabetic subjects. Diabetes 32:723-729, 1983 33. Itoh M, Hirooka Y, Nihei N: Response of plasma somatostatin-like immunoreactivity (SLI) to a 75 g oral glucose tolerance test in normal subjects and patients with impaired glucose tolerance. Acta Endocrinol 104:468-474, 1983 34. Hokfelt T, Johansson O, Ljungdahl A, Lundberg JM, Schultzberg M: Peptidergic neurones. Nature 284:515-521, 1980 35. Arimura A, Sato H, Dupont A, Nishi N, Schally AV: Somatostatin: Abundance of immunoreactive hormone in rat stomach and pancreas. Science 189:1007-1009, 1975 36. Sanders KM: Role of prostaglandins in regulating gastric motility. Am J Physiol 247:Glt7-GI26, 1984 37. Misiewicz JJ, Waller SL, Kiley N, Hotton EW: Effect of oral prostaglandin E1 on intestinal transit time in man. Lancet 1:648, 1969 38. Matuchansky C, Bernier JJ: Effect of prostaglandin E1 on glucose, water and electrolyte absorption in the human jejunum. Gastroenterology 64:1111-1115, 1973 39. Thompson DG, Wingate DL, Thomas M, Harrison D: Gastric emptying as a determinant of the oral glucose tolerance test. Gastroenterology 82:51-55, 1982 40. Schusdziarra V, Harris V, Arimura, A, Unger RH: Evidence for a role of splanchnic somatostatin in the homeostasis of ingested nutrients. Endocrinology 104:1705-1708, 1979 41. Malaisse-Lagae F, Carpentier JL, Patel YC, Malaisse WJ, Orci L: Pancreatic polypeptide: Possible role in the regulation of food intake in them mouse. Hypothesis. Experientia 33:915-917, 1977 42. Lin TM: Pancreatic polypeptide: Isolation, chemistry and biological function. In Gastrointestinal Hormones. GBJ Glass (ed). New York, Raven Press, 1980, pp 275-306 43. Penston JG, Johnston DA, Wormsley KG: The effects of rioprostil on gastric emptying and intragastric acidity. Hepato-gastroenterology 33:120-122, 1986 Digestive Diseases and Sciences, Vol. 35, No. 1 (January 1990)
Effect of Rioprostil, a Methylprostaglandin E 1 Analog, on Basal and Stimulated Plasma Pancreatic Hormone Levels in Man O. SEGERS, MD, and G. SOMERS, MD, PhD
The effect of rioprostil, a methylprostaglandin E 1 analog on circulating pancreatic hormones was evaluated in 13 healthy male subjects. Rioprostil administration, 300 tzg twice daily resulted in a significant decrease of fasting insulin, C-peptide, glucagon, and pancreatic polypeptide. No change in fasting plasma glucose or somatostatin levels was observed. An oral glucose tolerance test induced similar increments in plasma glucose concentration before and during treatment, but a delayed rise o f insulin and C-peptide levels occurred during the administration of the drug. On rioprostil, the glucose load no longer inhibited peripheral glucagon or somatostatin. Treatment with rioprostil remained without effect on mixed meal-induced changes in plasma glucose levels and concomitant increases in insulin, pancreatic polypeptide, and somatostatin levels. I t is concluded that in healthy individuals rioprostil influences the basal and glucose-induced levels Of glucagon, insulin, and somatostatin. In healthy men this effect did not, however, result in glucose intolerance. KEY WORDS: Prostaglandins; rioprostil; pancreatic hormones; glucose tolerance.
Synthetic prostaglandin (PG) analogs recently have been introduced in the treatment of peptic gastric and duodenal ulcers for their antisecretory (1-3) and cytoprotective (4-6) properties. During the past decade, however, several reports dealt with the effect of PG on the hormone secretion of the main pancreatic islet cells, namely the A, B, D, and PP cells, thereby influencing glucose homeostasis and possibly regulating the ingestion of nutrients. In man, PGE1 and PGE2 were described to cause inhibition of glucose-induced insulin secretion (7, 8). Likewise, all nonsteroidal antiinflammatory drugs except indomethacin, which are well-known Manuscript received January 30, 1989; revised manuscript received August 31, 1989; accepted September 15, 1989. From the Department of Internal Medecine, University Hospital AZ-VUB, Brussels, Belgium. Address for reprint requests: Dr. G. Somers, Department of Internal Medecine, Akademisch Ziekenhuis VUB, Laarbeeklaan 101, B1090 Brussels, Belgium.
20
PG synthesis inhibitors, increased basal and glucose-induced insulin levels in normal subjects (7, 9-14). Plasma glucagon levels were enhanced by PGE 1 infusion (15-17). This enhancement appeared to be an indirect phenomenon mediated either by endogenously released cathecholamines or by the interaction b e t w e e n the PG and the alphaadrenoreceptor. PG infusion has been described to stimulate gastric and pancreatic somatostatin release in vitro and in vivo (18, 19). Indomethacin apparently attenuated the postprandial somatostatin response in dogs (20). Only a few reports are available on the effects of prostaglandin analogs on pancreatic B-cell function in man. A variety of PGE~ and E2 analogs have been used in these studies. Enprostil, a PGE2 analog, caused a reduced rise in insulin after a standard meal without change in plasma glucose (21). Following administration of another PGE2 analog, eiDigestive Diseases and Sciences, Vol. 35, No. 1 (January 1990)
0163-2116/90/0100-0020506.00/09 1990PlenumPublishingCorporation
RIOPROSTIL AND PANCREATIC HORMONES ther orally, intraduodenally, or intravenously, serum insulin excursions in response to glucose were decreased (22). Misoprostol, a PGE1 analog; induced a significant e n h a n c e m e n t of plasma glucose levels and reduction of insulin output after an oral glucose load but did not change the insulin secretion after an intravenous glucose load (23). Reports on the effect of the PG analogs on the other pancreatic h o r m o n e s are scarce. Solely P G E 2 analogs have been described to reduce the p e a k p l a s m a levels of pancreatic polypeptide (PP) in r e s p o n s e to a meal (24). Postprandial p l a s m a somatostatin levels in man were not altered by either exogenous P G E 1 administration or depletion by indomethacin (25). The aim of our study was to analyze the acute and m o r e prolonged effects of oral administration of rioprostil, a PGE~ analog, on the fasting and postprandial p l a s m a glucose and on the concomitant changes in p l a s m a insulin, PP, somatostatin-like immunoreactivity (SLI), and glucagon levels in healthy individuals.
MATERIALS AND METHODS Study Protocol. The study protocol had the approval of the Ethical Committee of the University Hospital Brussels. In an open study 13 healthy male volunteers received 300 Ixg rioprostil orally, twice daily during one week. The mean age of the healthy subjects was 28 + 1 years (mean -+ SE, range 23-38 years). Body weight was within 5% of normal (26). The subjects did not take any drugs except for the test medication during the study period. A glucose load of 75 g was given in basal conditions, and after 24 hr and seven days of treatment with the methylprostaglandin E 1 analog rioprostil. A mixed meal test of 550 kcal (40% carbohydrate, 30% fat. and 30% protein) was given before and on day 6 of treatment. Meal intake time was kept below 7 min. Blood was drawn for biochemical analysis after a rest period of 15 min following the insertion of the intravenous sampling needle and was collected in chilled glass tubes containing EDTA and aprotinin (Trasylol, Bayer. Leverkfisen, FRG over a 150-min period. Blood was immediately centrifuged at 4~ and plasma was stored at - 2 0 ~ until assay. Materials. Rioprostil tablets (300 ~g) were prepared by Cilag, Schaffhausen, Switzerland. During the study, possible side or adverse effects were asked for and recorded daily. On the test day rioprostil was given 1 hr before oral glucose or mixed meal. Biochemical Analysis. Commercially available kits were used for the radioimmunological determinations of plasma insulin (Pharmacia. Uppsala. Sweden), C-peptide, and pancreatic glucagon (IRE, Fleurus. Belgium). PP and SLI were measured as previously described (27, 28). All samples of the same individual were determined in one Digestive Diseases and Sciences, Vol. 35, No. 1 (January 1990)
single radioimmunoassay. Plasma glucose was measured by glucose oxidase method. Statistical Methods. Results are expressed as mean SE. Significances of differences were estimated using an analysis of variance test (ANOVA). The significance level was set at 5%. Fasting plasma glucose and hormone levels as expressed in Figure 1 are the means of the basal values scored on the days of oral glucose tolerance test (OGTT) and/or mixed meal.
RESULTS Fasting Plasma Glucose and Hormone Levels. Figure 1 shows the fasting p l a s m a glucose and horm o n e levels before any drug intake and after one day and one w e e k of rioprostil treatment. The fasting p l a s m a glucose and S L I levels did not significantly change during the trial. A minor but significant decrease in fasting insulin, and C-peptide was o b s e r v e d during rioprostil administration. A more p r o n o u n c e d decline in fasting peripheral glucagon levels was found in the rioprostil period. M o r e o v e r , fasting glucagon levels w e r e significantly lower after 24 hr than after one w e e k of treatment with the drug. In contrast, fasting p l a s m a PP levels were only significantly d e c r e a s e d after one week of administration of the PG analog. Response to Oral Glucose. Plasma glucose excursions after the oral glucose load were not affected by a 24 hr t r e a t m e n t with rioprostil (Figure 2, upper panel). H o w e v e r , the induced rise in peripheral insulin levels a b o v e basal values was significantly lower after one day of prostaglandin treatment in c o m p a r i s o n to the increase in insulin concentrations before administration of the drug (Figure 2, lower panel). The integrated increment in p l a s m a insulin from the 30th to the 60th min after oral glucose averaged 1266 _+ 219 mU/liter on rioprostil compared to 1593 + 324 mU/liter before drug intake (P < 0.05). Again, after s e v e n days o f rioprostil, no difference in p l a s m a glucose increments could be o b s e r v e d in c o m p a r i s o n with the results obtained before t r e a t m e n t (Figure 3, u p p e r panel). H o w e v e r , in one of 13 patients a late and s y m p t o m a t i c hypoglycemia was noted after 150 min. After one week of rioprostil intake a significantly delayed rise of plasma insulin and C-peptide levels occurred (Figure 3, middle and lower panels) to levels that were slightly but significantly e n h a n c e d at 60 min after the ingestion of the c a r b o h y d r a t e . Peripheral glucagon levels w e r e inhibited b y the glucose load, After one w e e k of rioprostil h o w e v e r , the already lowered basal peripheral glucagon concentrations were
21
SEGERS AND SOMERS
8-
4 E E
3
LU
09 O O :D ._1 (3
2 1
o ._i 0
E
6"
z "i
4
co z
2
10 84
0
0
z O (,9
20 84
10-
5
B
D1
D6 7
01 .... =I DI
DI
B
D6 7
2~ ~
o
1
/// ///
I/i
iii
/~/ // /
o_
E ~
))j
--
DI
D6 7
120 -
~ 1J.I
,oo
~
80
n
60"
y/
0 B
B~
D6-7
~,/,'. B
. DI
40 D6-7
B
DI
D6-7
Fig 1. Fasting plasma glucose and pancreatic hormone levels before (B) and after 24 hr (D1) and six to seven days of rioprostil treatment (D6-7).
no longer significantly influenced by the glucose load (Figure 4, upper panel). In these healthy individuals, ingestion of 75 g of glucose resulted in an erratic change in peripheral SLI within 60 min after the glucose load and in a subsequent progressive decline of the circulating hormone levels, which OGTT
10'
0.05).
E u) o o D .J ~9
5
0
0
30
6'0
9'0
120
150
0
3'0
6'0
9'0
120
150
80 60 E i
(n z
40
20 0
TIME(min)
Fig 2. Peripheral plasma glucose (upper panel) and insulin levels (lower panel) during OGTT before (open symbols) and after 24 hr of rioprostil treatment (closed symbols).
22
became significantly lower than basal values from the 60th min after glucose intake. The SLI levels indeed progressively decreased to 62 + 7% of the basal value at 150 min after glucose intake. After seven days of rioprostil treatment, no significant reduction in peripheral SLI occurred (Figure 4, middle panel), the levels of SLI being significantly higher in comparison to the levels measured before drug treatment at 150 min and still averaging 94 + 7% of the corresponding basal concentration (P < Oral glucose did not induce any significant change in peripheral PP levels prior to or during rioprostil treatment (Figure 4. lower panel). Nevertheless in the one patient manifesting hypoglycemm after 150 min. a brisk increase in peripheral PP concentration was observed. Response to Mixed Meal. The responses of plasma glucose and the different hormones studied to ingestion o f a 550-kcal meal are shown in Figures 5 and 6. Six days of rioprostil treatment did not significantly influence the postprandial increments in plasma glucose, insulin, and C-peptide. somatostatin, or pancreatic polypeptide. Before and during rioprostil treatment, plasma glucose increased slightly to a comparable peak after 30 min (Figure 5, upper panel). Plasma insulin and C-peptide concentrations displayed a similar rise (Figure 5 middle and lower panel). Incidentally, 90 min after the ingestion of Digestive Diseases and Sciences Vol. 35. No. l (January 1990)
RIOPROSTIL AND PANCREATIC HORMONES OGTT
OGTT
~ E ~-
10'
100'
8
80' Z
60'
O
4-
0
40'
._1 (5
2"
o ..J o
20
6
O
O3
0 0
3'0
8'o
9'0
1~o ~5o
100
m ..,I
O3 Z
60 40
0
_--E
15
~,~
10
3'o
6'o
9'o
~o
1~o
E . El~,O . ~
9'0
12o 1~o
o
3'o
6'0
9'o
12o 1~o
0
30
150
8
~..
6-
r3
6'0
200
12
LU a.
3'0
20' 0
=~
o
20 84
80 E
0
100
r
4-
50
2"
0
0
o
3'o
6'o
9'o
1~o 1;o
TIME(mIn)
Fig 3. Peripheral plasma glucose (upper panel), insulin (middle panel), and C-peptide levels (lower panel) during OGTT before (open symbols) and after seven days of rioprostil administration (closed symbols).
the meal the plasma glucose and concomitant insulin and C-peptide levels were significantly higher on rioprostil. In basal conditions, the meal test induced a small but significant rise of plasma SLI over the fasting level from 30 to 120 min after the start of food ingestion (Figure 6, upper panel). During rioprostil administration a significant increase in SLI above fasting levels occurred only at 30 and 120 min after the start of the meal. However, the integrated release of SLI from 0 to 150 min was not significantly affected by the treatment with the drug (Figure 6, upper panel). The test meal also induced a biphasic rise in plasma PP levels. Again this response was not significantly altered by the treatment with the PG analog (Figure 6 lower panel). Glucagon was not measured during the meal test. Adverse Effects of Rioprostil. No adverse effects with 600 Ixg rioprostil daily were observed. Two Digestive Diseases and Sciences, Vol. 35, No. 1 (January 1990)
f
f
60 9'0 TIME(min)
120
150
Fig 4. Peripheral glucagon (upper panel), SLI (middle panel), and PP levels (lower panel) during OGTT before (open symbols) and after seven days of rioprostil treatment (closed symbols).
individuals mentioned a minor change in bowel habits with more loose stools. As already mentioned, in one subject a late hypoglycemia occurred after oral glucose load. DISCUSSION
The results of the present study suggest that rioprostil administration, 600 txg daily, to healthy volunteers influences the fasting and/or the OGTTinduced peripheral levels of insulin, glucagon, SLI, and PP. Although no effect of the drug on fasting glycemia and glucose tolerance was observed, basal plasma insulin and C-peptide levels were significantly lower during rioprostil treatment. Twentyfour hours after starting rioprostil administration, insulin secretion was decreased after an oral glucose load, which is consistent with already published data on the acute, direct effect of PG and its
23
SEGERS AND SOMERS MEAL TEST MEAL TEST
25 E
E
0,.
o0 o(D
20
...1
15
(5
3'0
e'o
9'0
z'o
8'0
9'o
,2o ,5o
3'0
6'0
9'0
120
120 400
60 300
50
E m ..J
40
,",
30"
13. 100
20"
z
10
0
0 0
3'0
6o
9o
LLI Q,.
O
Fig 6. Mixed meal-induced peripheral SLI (upper panel) and PP levels (lower panel) before (open symbols) and after six days of Rioprostil treatment (closed symbols).
4 2"
0
3'0
6'0
90
120
150
TIME(min)
Fig 5. Mixed meal-induced peripheral plasma glucose (upper panel), insulin (middle panel), and C-peptide levels (lower panel) before (open symbols) and after six days of rioprostil administration (closed symbols).
analogs on insulin secretion in man (21-23). It has to be stated that the dose of rioprostil used in this study was that recommended by the manufacturer for the treatment of peptic ulcer disease in man. No data on dose-response relationships of this compound on carbohydrate metabolism and pancreatic or intestinal hormone secretion are actually available. Nevertheless, in a recent study, administration of 300 Ixg of rioprostil, 25 min prior to an OGTT, was reported to remain without effect on basal insulin levels, but apparently caused a delayed increase in insulin levels without any changes in glucose tolerance (29). Part of the decline in insulin response observed after the administration of prostaglandin analogs was thought to be attributable to an inhibitory effect of the drugs on the
24
150
TIME(min)
8
I,Q,.
200
intestinal release of gastric inhibitory polypeptide (GIP), thus by abolishing its incretin effect (21, 22, 29). At the dose of 300 Ixg rioprostil indeed inhibited GIP secretion following an OGTT (29). Whether rioprostil administration affects other additional incretin factors such as glucagon-like peptide remains unknown (30). However, in light of the inhibitory action of prostaglandins demonstrated in vitro on B-cell function, a direct effect of the PG analog could well be at the origin of the changes in insulin output. This study further demonstrates that some effect of rioprostil on insulin excursions still persists after one week of treatment. During an OGTT, insulin and C-peptide secretion indeed displayed a slower increase, but the amplitude of the insulin response appeared to be unaffected. This delayed secretory pattern was probably the basis of the late hypoglycemic symptoms reported by one subject. The fasting pancreatic glucagon secretion was also significantly lowered during rioprostil treatment. This seems at variance with some reported data (15, 16). The reasons for the different findings could be related to the different types and routes of administration of the PG used in the different studies or to a specific action of this particular PG analog. Indeed different effects of PGE 1 and PGE2 on hormone secretions have been demonstrated Digestive Diseases and Sciences, Vol. 35, No. 1 (January 1990)
RIOPROSTIL AND PANCREATIC HORMONES
(29). Whether the observed decrease in peripheral glucagon levels results from a direct effect of the prostaglandin analog on the A-cell secretion cannot be answered by this study. However, the concomitant decline of plasma glucagon could balance the inhibited insulin output during rioprostil treatment, thereby contributing to the lack of any apparent disturbances in plasma glucose concentrations. Moreover, peripheral glucagon following an oral glucose load remained at the level of this decreased fasting glucagon concentration. The basal and stimulated plasma SLI did not change during the study. Similar findings were reported during misoprostol treatment, another PGE~ analog (25). Oral glucose has been reported to either stimulate or inhibit peripheral SLI levels (31-33). In our experience, no significant rise in circulating SLI occurred after a glucose load in healthy individuals. On the contrary, a significant decline was observed. This inhibitory effect of glucose on peripheral SLI levels appeared to be blunted during rioprostil treatment. Indeed, higher SLI levels were observed from 120 min on after the glucose intake, concomitant with the decline of the stimulated insulin levels. The interpretation of this data is difficult and may not point to any specific alterations in pancreatic D-cell function. First, only a minor, although significant, difference induced by rioprostil could be detected. Second, somatostatin is not only released into the circulation by the pancreatic D cells, but also by somatostatinproducing cells of the gastrointestinal tract and is also located in intestinal nerve fibers (34, 35). One could speculate, however, that this lack of decline in SLI somehow could play a role in the maintenance of normal glucose excursions in the presence of altered peripheral and presumably portal insulin and glucagon concentrations. The fasting PP secretion was decreased only after prolonged PG treatment. The mixed meal test after six days of treatment could not demonstrate any difference in secretory reaction of the pancreatic islet PP cell, This is in contrast with the findings with a PGE2 analog in duodenal ulcer patients in whom an inhibitory effect on stimulated PP cell secretion was observed (24). It can be advanced that part of eventual metabolic effects of prostaglandin analogs could be due to the influence of these drugs on gastric emptying, intestinal mortility, glucose absorption, or even on gastrin release (36-38). Alterations of gastric emptying apparently primarily influence postprandial Digestive Diseases and Sciences, Vol. 35, No. 1 (January 1990)
glucose levels (39). A more rapid presentation of carbohydrates at the level of the duodenum indeed can result in higher blood glucose concentrations (39). In our study, no such effect could be observed. Furthermore, no influence of rioprostil on basal and stimulated gastrin levels could be demonstrated for the rioprostil dose of 600 ~g daily (3). Likewise, rioprostil administration had no effect on the mealinduced secretion of PP and somatostatin, pancreatic hormones that apparently play a role in the regulation of either food intake, absorption or digestion (40-42). Nevertheless, it has to be mentioned that at the dose used in our study a small but significant increase in gastric emptying has been reported (43). Oral administration of rioprostil for one week was well tolerated by the healthy volunteers and no serious side effects on bowel habits were noted at the dose used. In conclusion, rioprostil treatment lowers insulin, C-peptide, PP, and glucagon levels; these alterations did not, however, induce glucose intolerance in healthy volunteers. The delayed rise of peripheral insulin levels during a glucose load could, however, result in more pronounced changes in glucose metabolism in either non-insulin-dependent diabetic patients or in individuals with glucose intolerance and an already altered insulin output. These findings suggest that careful attention should be given to such patients treated with this PG analog. ACKNOWLEDGMENTS
The authors thank Gaby Schoonjans for his technical assistance and Nadine Van Slycke for helping with the manuscript. REFERENCES 1. Akdamar K, Agrawal NM, Ertran A: Inhibition of nocturnal gastric secretion in normal human volunteers by misoprostol: A synthetic prostaglandin El, methyl ester analog. Am J Gastroentoerol 77:902-904, 1982 2. Wilson DE, Quadros E, Rajapaksa T, Adams A, Noar M: Effects of misoprostol on gastric acid and mucus secretion in man. Dig Dis Sci 31(suppl):1265-1295, 1986 3. Schulte K, Singer MV, Eysselein V, Demol P, Goebell H: Effect of rioprostil, a synthetic prostaglandin E 1, on meal stimulated gastric acid secretion and plasma gastrin levels in humans. Digestion 36:162-167, 1987 4. Robert A: Cytoprotection by prostaglandins. Gastroenterology 77:764-767, 1979 5. Detweiler MK, Harrison CA, Rollins DE, Tolman KG, MacCormick GM, Simon DM: Effects of riopostil on aspirin induced gastrointestinal mucosal changes in normal volunteers. Gastroenterology 86: P1062, 1984 (abstracts)
25
SEGERS AND SOMERS 6. Agrawal NM, Godiwala T, Arimura A, Dajani EZ: Cytoprotection by a synthetic prostaglandin against ethanol-induced gastric mucosal damage. A double blind endoscopic study in human subjects. Gastrointest Endocrinol 32:67-70, 1986 7. Robertson RP, Chen M: A role for prostaglandin E in defective insulin secretion and carbohydrate intolerance in diabetes mellitus. J Clin Invest 60:747-753, 1977 8. Giugliano D, Torella R: Prostaglandin E1 inhibits glucoseinduced insulin secretion in man. Prostaglandin Med 1:165166, 1978 9. Field JB, Boyle C, Remer A: Effects of salicylate infusion on plasma insulitis and glucose tolerance in healthy persons and mild diabetics. Lancet 1:1191-1194, 1978 10. Micossi R, Pontiroli AE, Baron SH, Tamayo RC, Lengel F, Bevilacqua M, Raggi U, Norbiato G, Foa PR: Aspirin stimulates insulin and glucagon secretion and increases glucose tolerance in normal and diabetic subjects. Diabetes 27:1196-1204, 1978 11. Chen M, Roberston RP: Effects of prostaglandin synthesis inhibitors on human insulin secretion and carbohydrate tolerance. Prostaglandins 18:557-567, 1979 12. Syv/ilahti EKG: The effect ofindomethacin on serum growth hormone, immunoreactive insulin, and blood glucose levels of young adult males. Int J Clin Pharmacol 10:111-116, 1974 13. Widstr6m A: Influence of indomethacin on glucose-induced insulin response in normal man-role of prostaglandins in the rapid insulin release? Horm Metab Res 9:172-175, 1977 14. Topoi E, Brodows RG: Effects of indomethacin on acute insulin release in man. Diabetes 29:379-382, 1980 15. Giugliano D, Torella R, Sgambato S, D'Onofrio F: Effect of ct and [3 adrenergic inhibition and somatostatin on plasma glucose, free fatty acid, insulin, glucagon, growth hormone responses to prostaglandin E 2 in man. J Clin Endocrinol Metab 48:302-308, 1979 16. Giugliano D, Torella R, D'Onofrio F: Prostaglandins and the alpha-cell. Prostaglandins Med 6:283-297, 1981 17. Sacca L, Perez G: Influence of prostaglandins on plasma glucagon levels in the rat. Metabolism 25:127-130, 1976 18. Safouri B, Weir GC, Bitas KN, Makhlouf GM: Gastrin and somatostatin secretions by perfused rat stromach: Functional linkage of antral peptides. Am J Physiol 238:495-501, 1980 19. Schusdziarra V, Rouiller D, Jaffe BM, Harris V, Unger RH: Effect of exogenous and endogenous prostaglandin E upon gastric endocrine function in dogs. Endocrinology 106:16201627, 1980 20. Schusdziarra V, Rouiller D, Harris V, Wasada T, Unger RH: Effect of prostaglandin E2 upon release of pancreatic somatostatin-like immunoreactivity. Life Sci 28:2099-2102, 1981 21. Nicholl CG, Carolan G, Sevelius H, Bloom SR: Enprostil reduces the post-prandial rises in insulin and glucosedependent insulinotropic peptide and delays small intestinal transit time. Diabetologia 29:575A, 1986 22. Konturek SJ, Mikos Em, Krol R, Wierzbicki Z, Dobrzanska M: Effect of methylated prostaglandin E2 analogue on insulin secretion in man. Prostaglandins 15:591-602, 1978 23. Ebert R, Hielscher A, Creutzfeldt W: Disruption of the enteroinsular axis by misoprostol. Eur J Clin Invest 17:A3, 1987 24. Petersen W, Feldman M, Taylor I, Brenner M: The effect of 15(R)-15-methyl prostaglandin E 2 on meal-stimulated gastric acid secretion, serum gastrin, and pancreatic polypeptide in duodenal ulcer patients. Dig Dis Sci 24:381-384, 1979
26
25. Lucey MR, Clark ML, Fairclough PD, Wass JAM, Webb J, Davidson K, Rees LM, Dawson AM: The effect of a prostaglandin agonist and antagonist on postprandial plasma somatostatin and gastrin in man. Gastroenterology 86:1167, 1984 26. The Metropolitan Life Insurance Company. Statistical Bulletin 45, 1959 27. Pipeleers DG, Pipeleers-Marichal MA: A method for the purification of single A, B, and D cells and for the isolation of coupled cells from isolated islets. Diabetologia 20:654663, 1981 28. Somers G, Pipeleers-Marichal M, Gepts W, Pipeleers D: A case of duodenal somatostatinoma: diagnostic usefulness of Calcium-pentagastrin test. Gastroenterology 85:1192-1198, 1983 29. Nauck M, Harter S, Ebert R, Creutzfeldt W: Effects of four orally administered analogues of prostaglandin E1 and E2 on glucose tolerance and on the secretion of pancreatic and gastrointestinal hormones in man. Eur J Clin Invest 19:298305, 1989 30. Kreymann B, Ghatei MA, Williams G, Bloom SR: Glucagon-like peptide 1 7-36: A physiological incretin in man. Lancet 2:1300-1304, 1987 31. Pimstone B, Berelowitz M, Kranold D, Shapiro B, Kronhelm S: Somatostatin-like Immunoreactivity in human and rat serum. Metabolism 27(Suppll):1145-1149, 1978 32. Conlon JM, McCulloch AJ, Alberti KGMM: Circulating somatostatin concentrations in healthy and non-insulin-dependent (type II) diabetic subjects. Diabetes 32:723-729, 1983 33. Itoh M, Hirooka Y, Nihei N: Response of plasma somatostatin-like immunoreactivity (SLI) to a 75 g oral glucose tolerance test in normal subjects and patients with impaired glucose tolerance. Acta Endocrinol 104:468-474, 1983 34. Hokfelt T, Johansson O, Ljungdahl A, Lundberg JM, Schultzberg M: Peptidergic neurones. Nature 284:515-521, 1980 35. Arimura A, Sato H, Dupont A, Nishi N, Schally AV: Somatostatin: Abundance of immunoreactive hormone in rat stomach and pancreas. Science 189:1007-1009, 1975 36. Sanders KM: Role of prostaglandins in regulating gastric motility. Am J Physiol 247:Glt7-GI26, 1984 37. Misiewicz JJ, Waller SL, Kiley N, Hotton EW: Effect of oral prostaglandin E1 on intestinal transit time in man. Lancet 1:648, 1969 38. Matuchansky C, Bernier JJ: Effect of prostaglandin E1 on glucose, water and electrolyte absorption in the human jejunum. Gastroenterology 64:1111-1115, 1973 39. Thompson DG, Wingate DL, Thomas M, Harrison D: Gastric emptying as a determinant of the oral glucose tolerance test. Gastroenterology 82:51-55, 1982 40. Schusdziarra V, Harris V, Arimura, A, Unger RH: Evidence for a role of splanchnic somatostatin in the homeostasis of ingested nutrients. Endocrinology 104:1705-1708, 1979 41. Malaisse-Lagae F, Carpentier JL, Patel YC, Malaisse WJ, Orci L: Pancreatic polypeptide: Possible role in the regulation of food intake in them mouse. Hypothesis. Experientia 33:915-917, 1977 42. Lin TM: Pancreatic polypeptide: Isolation, chemistry and biological function. In Gastrointestinal Hormones. GBJ Glass (ed). New York, Raven Press, 1980, pp 275-306 43. Penston JG, Johnston DA, Wormsley KG: The effects of rioprostil on gastric emptying and intragastric acidity. Hepato-gastroenterology 33:120-122, 1986 Digestive Diseases and Sciences, Vol. 35, No. 1 (January 1990)
