Mechanism of the Antidiarrheal Effect of Loperamide

GASTROENTEROLOGY 1984;86;1475-80

Mechanism of the Antidiarrheal Effect of Loperamide LA WRENCE R. SCHILLER, CAROL A. SANTA ANA, STEPHEN G, MORAWSKI, and JOHN S, FORDTRAN Departments of Internal Medicine, Baylor University Medical Center, Dallas, Texas, and Veterans Administration Medical Center, Dallas, Texas

To determine whether the antidiarrheal effect of loperamide is due to an effect on intestinal motor function or to an acceleration of the rate of absorption by the intestine (as has been suggested recently), we studied absorption during experimental diarrhea produced by the rapid intragastric infusion of electrolyte solution, In studies in which a 2700-ml bolus of electrolyte solution was infused into the stomach over 90 min, loperamide delayed the appearance of rectal effluent in each of 5 subjects and decreased the volume of rectal effluent from 1090 ± 118 to 770 ± 73 ml (p = 0.05). When intragastric infusion was continued for 5 h, producing steadystate total gut perfusion, the volume of effluent produced per unit time and the concentration of a nonabsorbable polyethylene glycol marker in rectal effluent was not different with or without loperamide, indicating that loperamide did not alter the rate of absorption by intestinal mucosal cells. Loperamide also had no effect during steady-state perfusion when absorption rates were reduced by intravenous infusion of vasoactive intestinal polypeptide. Loperamide did substantially increase the intraluminal volume of the total gut, from 985 ± 131 to 1764 ± 195 ml (p < 0.02). These results suggest that loperamide exerts its antidiarrheal effect by a change in the motor function of the intestine, which results in increased capacitance of the gut and a delay in the passage of fluid through the intestine. Received September 1, 1983. Accepted December 28, 1983. Address requests for reprints to: Lawrence R. Schiller, M.D., Department of Internal Medicine, Baylor University Medical Center, 3500 Gaston Avenue, Dallas, Texas 75246. This study was supported by U.S. Public Health Service Grant l-R01-AM26794 from the National Institute of Arthritis, Metabolism, and Digestive Diseases, and grants from Janssen Pharmaceutica, New Brunswick, New Jersey, and the Southwestern Medical Foundation's Abbie K. Dreyfuss Fund, Dallas, Texas. The authors thank Mrs. Bobbie Bullard for her expert secretarial assistance in preparation of this manuscript. © 1984 by the American Gastroenterological Association 0016-5085/84/$3.00

This change in motor function, rather than a change in the rate of absorption by intestinal mucosal cells, is responsible for the antidiarrhea'l effect of loperamide in our experimfJntal diarrhea model. There are at least two ways by which a drug might increase net absorption by the gut, and thus have an antidiarrheal action: (a) the drug might stimulate the rate of absorption by intestinal epithelial cells, or (b) the drug might slow the movement of fluid through the gut, allowing more time for absorption to occur. Either of these actions could result in greater overall absorption and a decrease in stool volume. In previous studies we found that the second of these actions accounts for the antidiarrheal activity of codeine (1). In this study we examined the mechanisms of action of a differ!=)nt antidiarrheal drug, loperamide. Like codeine, loperamide is effective in the treatment of patients with diarrheal diseases of a variety of etiologies (2-10). Loperamide, however, has certain pharmacologic properties that diffElf from codeine and which may fllter its mechanism of action. Unlike ~odeine and other more traditional opiate antidiarrheal drugs, 10perall1ide can bind to and inhibit calmodulin (11,12)' a protein thought to be an important regulator of intestinal ion transport (13). Loperamide also has a different distribution within the body than other opiates. Whereas drugs like morphine and codeine penetrate the central nervous system and interact with central opiate receptors to modify intestinal motor function (1416), only small concentrations of loperamide reach the central nervous system, and thus its effects are thought to be due mainly to interactions with peripheral opiate receptors (16,17). These opiate receptors may be of the mu-subtype like codeine (1), or may be of other subtypes such as 8- or K-receptors; Abbreviations used in this paper: AMP, adenosine monophosphate; PEG, polyethylene glycol; VIP, vasoactive intestinal polypeptide.

1476

SCHILLER ET AL.

this issue has not been well studied. Some investigators have even speculated that the action of loperamide is due to a non-opiate-mediated process (18). Whether due to these pharmacologic differences or not, a variety of evidence has suggested that loperamide has specific antisecretory or proabsorptive effects on intestinal mucosa that are mainly responsible for its ant~diarrheal effect. Animal studies in vivo indicated that loperamide inhibits fluid secretion stimulated by prostaglandins, cholera toxin, or bisacodyl (19-22). Preliminary studies in humans using segmental jejunal perfusion also indicated that 10peramide may stimulate absorption rates or may inhibit secretion rates under some circumstances (23,24). Studies in vitro have differed in details but suggested that loperamide reduces the serosa to mucosa flux of chloride (25-:P) or increases the mucosa to serosa flux of sodium, or both (22,25,27). These observations are compatible with inhibition of the rate of secretion or stimulation of the rate of absorption by mucosal cells. The issue of whether or not loperamide has proabsorptive or antisecretory effects that contribute to its antidiarrheal action is of considerable theoretical and practical interest and importance. Many diarrheal illnessep' are the result of reduced rate of absorptiop or an excessive rate of secretion by intestinal mucosal cells ("secretory diarrhea"). Having a specific agent that is capable of reversing these defects would probably improve therapy. To determine whether or not therapeutic doses of loperamide have proabsorptive or anti secretory effects in humans, we studied the effect of loperamide on experimental diarrhea induced by the rapid intragastric infusion of electrolyte solution in normal subjects, with and without intravenous infusion of the intestinal secretagogue, vasoactive intestinal polypeptide (VIP). Using this technique we were able to quantitate the antidiarrheal effect of loperamide, and evaluate its mechanism of action with regard to its influence on absorption, secretion, and motility.

Metqods Subjects Six healthy men aged 25-28 yr (mean, 26 ± 0.5 yr) and weighing from 70 to 79 kg (mean, 76.3 ± 1.4 kg) participated in these studies after giving written informed consent. The consent forms and study protocols were approved by an Institutional Review Board.

Subject Preparation To empty the gut of fecal debris, each subject rapidly ingested 5 L of a poorly absorbable mannitolsodium sulfate solution (Golytely, Braintree Laboratories,

GASTROENTEROLOGY Vol. 86. No. 6

Inc., Braintree, Mass.) (28) during the evening before each experiment. Immediately afterward, each subject rapidly ingested 2 L of isotonic saline in order to flush out any remaining poorly absorbable solution. Previous studies have inqicated that this preparation results in complete removal of fecal debris from the intestine (1,28).

. Drugs Loperamide hydrochloride (Janssen Pharmaceutica, New Brunswick, N.J.) was given in a total dose of 18 mg: 8 mg 6 h before starting the experiment and 10 mg 2 h before starting the experiment. This total dose is approximately four times the average daily dose required to control chronic diarrhea (4). In some experiments, blood was drawn for measurement of plasma loperamide levels. Before control experiments, subjects ingested a similar number of placebo capsules at the same times. During some experiments, porcine VIP (Gastrointestinal Hormone Research Unit, Karolinska Institutet, Stockholm, Sweden) was infused intravenously at a rate of 300 pmollkg . h. This dose would be expected to substantially reduce water and electrolyte absorption in humans (29).

Experimental Diarrhea Produced by Intragastric Infusion of Balanced Electrolyte Solution After preparation and an overnight fast, each subject swallowed a single-lumen, mercury-weighted polyvinyl tube so that the tip of the tube was located in the stomach. A rectal tube was also placed at this time with its tip located 17 cm from the anal verge. Balanced electrolyte solution (NaCI 100 mM, KCI 4 mM, NaHC0 3 40 mM) containing polyethylene glycol (PEG, mol wt 3000-3700) 2 giL as a nonabsorbable marker was infused into the stomach at a rate of 30 mllmin. Rectal effluent was collected in 15-min samples via the rectal tube. Infusion was continued for either 1.5 or 5 h ("Bolus Study" and "Total Gut Perfusion Study," respectively; see Results).

Analysis of Samples, Calculations, and Statistics Effluent volume was measured in graduated cylinders and samples were analyzed for PEG concentration by the method of Hyden (30). During steady-state perfusion, measurement of PEG concentration in rectal effluent allowed calculation of the effluent flow rate as a check on the accuracy of the measured flow of rectal effluent. Effluent flow rates were calculated by the following formula: EFR

=

IR x [PEG];nfu,ate/[PEG]effluento

where EFR is the effluent flow rate and IR is the infusion rate. During steady-state perfusion, PEG measurements also allowed us to calculate total gut volume (TGV) by the following formula: TGV

=

amount of PEG in gutl[PEG] in gut.

June 1984

ANTIDIARRHEAL EFFECT OF LOPERAMIDE

The amount of PEG in the gut at any point in time was the difference between the amount infused into the stomach and the amount recovered by the rectal tube. [PEG] in the gut was assumed to be the mean of the [PEG] infused and the [PEG] recovered. Plasma samples were assayed for loperamide by radioimmunoassay (31) at Janssen Pharmaceutica, Beerse, Belgium. Means, standard errors of the mean (SEM), and p values by paired t-test were calculated by standard techniques (32). Values of p :s 0.05 were taken to be statistically significant.

Results Effect of Loperamide on Effluent Output in Response to Infusion of a Bolus of Fluid

The effectiveness of 18 mg of loperamide in reducing effluent output was assessed by infusing a 2700-ml bolus of fluid into the intestine over a 90min period and measuring effluent output after administration of loperamide or placebo on separate test days. When a bolus of fluid is introduced into the intestine, changes both in the rate of absorption by intestinal mucosal cells and in the time available for absorption (produced by motility changes) could affect effluent output (1). To assess the duration of action, a second bolus of fluid was infused 4 h after the first. Cumulative effluent volume during this experiment is shown in Figure 1. Output in response to the first bolus, the second bolus, or both together was significantly less after loperamide administration, indicating that an antidiarrheal effect was present during both bolus infusion periods. In every subject, the first appearance of rectal effluent was delayed by loperamide as compared with placebo. Effluent PEG

INTRAGASTRIC INFUSION

2.400

wr-

~i'5

2.000

c5~

~i:i= 1,600

>...J

E

~f'!- 1,200

5~

::;;'" ::>LL

uo

800 400 3

4

5

TIME (Hours)

6

7

8

Figure 1. Antidiarrheal effectiveness and duration of action of loperamide. Cumulative volume of rectal effluent is plotted versus time. A balanced electrolyte solution (2700 ml) was infused as a bolus from 0 to 1.5 hand again from 4 to 5.5 h. Effluent output was delayed in each subject, and significantly reduced for the group as a whole by loperamide (first bolus, p = 0.05; second bolus, p < 0.05; both boluses, p < 0.005).

1477

Table 1. Mean (± SE) Polyethylene Glycol Concentrations in Rectal Effluent During and After Bolus Infusion of 2700 ml of Balanced Electrolyte Solution in Five Subjects Placebo Loperamide p Value

At 2 h (giL)

Over 4 h (giL)

3.65 ± 0.26 3.57 ± 0.45 NS

3.11 ± 0.17 4.00 ± 0.24 <0.01

NS indicates that p value is not statistically significant (p > 0.05).

marker concentrations 2 h after the first bolus, a time at which a significant difference in volume output was present, were not significantly different between placebo and loperamide test days (Table 1), suggesting but not proving that a delay in the passage of fluid through the intestine rather than an increase in the rate of absorption was the primary reason for less effluent output with loperamide.· However, for the entire 4-h period after the first bolus, PEG marker concentrations were significantly higher with loperamide (Table 1), indicating an increase in net absorption with loperamide. This could be due either to an enhanced rate of absorption by mucosal cells or to prolonged contact of luminal fluid with intestinal mucosa. In six experiments, PEG was omitted from the second bolus of balanced electrolyte solution in order to measure PEG recovery from the first bolus. In those studies, PEG recovery averaged 102% ± 3%. Effect of Loperamide During Steady-State Perfusion of the Whole Gut

A bolus study cannot differentiate between effects on the rate of absorption and effects on motility because in such a study both factors can influence the output of rectal effluent. On the other hand, steady-state perfusion of the whole gut eliminates any effect of motility on effluent output because during the steady state all parts of the gut are continuously exposed to the infused fluid (1). Thus it is impossible to increase net absorption by delaying the flow of fluid and thus increasing the time available for absorption; the time available per unit area of mucosa is already maximal. Results of steady-state whole gut perfusion are shown in the left panels of Figure 2, In every subject, loperamide delayed the appearance of rectal effluent. However, once a steady state was reached (during the last 2 h, when volume output per unit time and PEG concentrations were stable), there was no difference between loperamide and placebo, either in effluent PEG concentrations (Figure 2, left lower panel) or in the measured rate of flow of rectal effluent (Table 2). Flow rates calculated from effluent

1478

SCHILLER ET AL.

"'z

5

~~~

4

>,,-

~~ -'w ~o:

=>"00

"'z wO a.f=

3

PLACEBO

\

2

I

I

I

/

~ 'LOPERAMIDE

-'0

bg

1

t

F~,

I

"

I

1,000

!~OPERAMli~

WITH VIP

1*

lOPERAMIOE...........

---+*

p--r--

1

3

5

2-2~

~PLACEBO

o

3

4

5

4

TIME AFTER STARTING INFUSION

(Hours)

f

LOPERAMIDE

,I~

fN"5)

(N=5)

002345012345 TIME (Hours)

Figure 2. Effect of loperamide on cumulative volume of rectal effluent and effluent polyethylene glycol (PEG) concentrations during total gut perfusion with and without simultaneous intravenous infusion of vasoactive intestinal polypeptide (VIP). Balanced electrolyte solution was infused intragastrically from 0 to 5 h. Loperamide delayed the initial appearance of rectal effluent but had no significant effect on the rate of flow of effluent or on effluent PEG concentrations.

PEG concentrations were very similar to measured flow rates (r = 0.93) and also were not significantly different with loperamide. These data indicate that loperamide had no effect on the rate of absorption by the intestinal mucosa. When VIP was infused intravenously, effluent volumes were increased and effluent PEG concentrations were reduced as compared with the test day when VIP was not infused (Figure 2 and Table 2), indicating that absorption rates were markedly reduced. Loperamide, however, did not effect the rate of flow of rectal effluent or effluent PEG concentrations (Figure 2, right lower panel), indicating that it did not reverse the VIP-induced reduction in absorption rate. Loperamide caused an almost twofold increase in calculated total gut volumes during steady-state whole gut perfusion, from 985 ± 131 to 1764 ± 195 ml (p < 0.02) (Figure 3, left panel). Simultaneous intravenous infusion of VIP caused no significant difference in total gut volume with or without loperamide (Figure 3, right panel). Plasma loperamide levels averaged 3.73 ± 0.33 ng/ Table 2. Mean (±SE) Rate of Flow of Rectal Effluent During the Fourth and Fifth Hours of Total Gut Perfusion at an Infusion Rate of 1800 rnllh

Placebo Loperamide P Value

ml

2,000

PLACEBO

4

~~22

diXl~RAMIDE

~r:i11't\

~:n,

~~:i 3 "-z

\r!t~X

.Cf

~~ LOPERAMIDE

I

0 5

PLACEBO

rrl 1'0'111

WITHOUT VIP

¥~ !Jrrl

6

33

0,,-

WITH VIP

WITHOUT VIP

7 wf-

GASTROENTEROLOGY Vol. 86, No. 6

Without VIP

With VIP

(mI/h)

(mllh)

914 ± 58 894 ± 117 NS

1490 ± 84 1460 ± 66 NS

NS indicates that p value is not statistically significant (p > 0.05).

Figure 3. Calculated total gut volumes during the last 2 h of steady-state whole gut perfusion with and without intravenous infusion of vasoactive intestinal polypeptide (VIP). As indicated by the asterisks, loperamide significantly increased total gut volume throughout this interval as compared with placebo (p < 0.05). Vasoactive intestinal polypeptide had no substantial effect on total gut volume by itself.

ml at the start of whole gut perfusion and 3.11 ± 0.14 ng/ml at the conclusion of the perfusion period. These values are slightly more than four times the peak loperamide concentrations achieved after a single 4-mg dose (31). Infusion of VIP did not significantly affect loperamide levels when measured at the conclusion of the perfusion period.

Discussion Loperamide had three striking effects in our studies. First, loperamide delayed the time of first appearance of rectal effluent in each subject in both the bolus and steady-state studies, indicating an effect on the movement of fluid through the gut. Second, loperamide significantly reduced the volume of rectal effluent produced during the bolus study, indicating that this dose of loperamide had a clear-cut antidiarrheal effect in this experimental diarrhea model. Third, loperamide significantly increased calculated total gut volume during total gut perfusion. Taken together, these results suggest that loperamide increases the capacitance of the gut and delays the movement of fluid through the intestine, allowing more time for absorption of fluid by the intestinal mucosa. When the effects of changes in motor function on net absorption were eliminated by continuing intragastric infusion so that the extra capacitance of the intestine was filled and the entire mucosa was continuously exposed to fluid during steady-state perfusion of the whole gut, loperamide had no effect on the flow rate or [PEG] of rectal effluent. This result indicated that loperamide did not have a demonstrable proabsorptive effect. This negative result was not due to using an insufficient dose of loperamide, as the dose used was more than four times the average daily dose needed to treat chronic diarrhea (4) and produced blood levels well within the therapeutic range throughout the perfusion period. Moreover,

June 1984

the same dose produced a clear-cut reduction in rectal effluent volume in the bolus experiments. This negative result was also not due to exceeding the drug's duration of action, as the steady-state experiments were carried out over a time period when the drug was still having an antidiarrheal effect as assessed by the bolus studies. Loperamide also had no observable antisecretory effect when tested against the intestinal secretagogue, VIP. Vasoactive intestinal polypeptide is thought to inhibit cation absorption and stimulate anion secretion by intestinal mucosal cells by means of increased intracellular cyclic adenosine monophosphate (AMP) levels (33). The failure of loperamide to reverse the decrease in absorption rate produced by VIP suggests that loperamide would not have a specific antisecretory effect against the wide range of toxins and secretagogues that produce diarrhea via cyclic AMP. These results are different than those obtained from several in vivo perfusion studies in humans and in rats under basal (20,24) or secretagoguestimulated (20,22,23) states and are different than those to be expected on the basis of in vitro studies (18,22,25-27). The reasons for this discrepancy remain unknown. Differences in methods, species, loperamide dose, and the type of secretagogue may be important. For instance, it is conceivable that some reported effects of loperamide, such as changes in paracellular permeability (34) or changes in intracellular calmodulin action (11-13), might be more readily observable under some experimental conditions than others. However, under the conditions of our experiments, loperamide had no effect on absorption rates by the gut as a whole, suggesting either that these mechanisms are not active in normal or VIP-stimulated human intestine in vivo, or that a therapeutic dose of loperamide does not affect them. It is possible that loperamide might have antisecretory effects against other secretagogues such as prostaglandins, but this would be surprising because both VIP and prostaglandins are thought to work via cyclic AMP. Our results suggest that the antidiarrheal effect of loperamide in various other human studies (8-10) with secretory states is probably due to an effect on gastrointestinal motor function. Our results with loperamide are analogous to those previously obtained with codeine (1). This similarity of action is particularly interesting because of differences in the distribution of loperamide and traditional opiate drugs within the body and alleged differences in the pharmacology of these agents (see introduction). In addition to having a direct effect on the gut, traditional opiates enter the central nervous system where even small amounts of drugs are able to produce substantial anti propulsive

ANTIDIARRHEAL EFFECT OF LOPERAMIDE

1479

(and antidiarrheal) effects (14-16). In animal studies, these central nervous system effects appear to be mediated by the vagus nerves (15) by means of pathways not involving opioid peptides in the periphery (16). On the other hand, therapeutic doses of loperamide do not substantially penetrate the central nervous system (17) and, consequently, have a predominant effect peripherally. Thus, whether mediated by both central and peripheral mechanisms (e.g., traditional opiate drugs) or exclusively by peripheral mechanisms (e.g., loperamide) and regardless of alleged differences in pharmacologic properties, the action of both traditional opiate antidiarrheal agents and loperamide appears to be similar-an inhibition of the propulsion of fluid through the gut.

References 1. Schiller LR, Davis GR, Santa Ana CA, Morawski SG, Fordtran JS. Studies of the mechanism of the antidiarrheal effect of codeine. J Clin Invest 1982;70:999-1008. 2. Galambos JT, Hersh T, Schroder S, Wenger J. Loperamide: a new antidiarrheal agent in the treatment of chronic diarrhea. Gastroenterology 1976;70:1026-9. 3. Shee CD, Pounder RE. Loperamide, diphenoxylate and codeine phosphate in chronic diarrhea. Br Med J 1980;1:524. 4. Palmer KR, Corbett CL, Holdsworth CD. Double-blind crossover study comparing loperamide, codeine and diphenoxylate in the treatment of chronic diarrhea. Gastroenterology 1980;79:1272-5. 5. Bergman L, Djarv 1. A comparative study of loperamide and diphenoxylate in the treatment of chronic diarrhoea caused by intestinal resection. Ann Clin Res 1981;13:402-5. 6. Wille-Jorgensen P, Gudmand-Hoyer E, Skovbjerg H, Andersen B. Diarrhoea following jejuno-ileostomy for morbid obesity. A randomized trial of loperamide and diphenoxylate. Acta Chir Scand 1982;148:157-8. 7. Sandhu BK, Tripp JH, Milia PI. Harries JT. Loperamide in severe protracted diarrhoea. Arch Dis Child 1983;58:39-43. 8. Yamashiro Y, Yamamoto K, Sato M. Loperamide therapy in a child with vipoma-associated diarrhoea. Lancet 1982;i:1413. 9. Karim SMM, Adaikan PG. The effect of loperamide on prostaglandin induced diarrhoea in rat and man. Prostaglandins 1977;13:321-31. 10. Lange AP, Secher NJ, Amery W. Prostaglandin-induced diarrhoea treated with loperamide or diphenoxylate. A doubleblind study. Acta Med Scand 1977;202:449-54. 11. Merritt JE, Brown BL, Tomlinson S. Loperamide and calmodulin. Lancet 1982;i:283. 12. Zavecz JH, Jackson TE, Limp GL, Yellin TO. Relationship between anti-diarrheal activity and binding to calmodulin. Eur J PharmacoI1982;78:375-7. 13. Ilundain A, Naftalin RJ. Role of Ca 2 + -dependent regulator protein in intestinal secretion. Nature 1979;279:446-8. 14. Parolaro D, Sala M, Gori E. Effect of intracerebroventricular administration of morphine upon intestinal motility in rat and its antagonism with naloxone. Eur J Pharmacol 1977;46: 329-38. 15. Stewart n, Weisbrodt NW, Burks TF. Central and peripheral actions of morphine on intestinal transit. J Pharm Exp Ther 1978;205:547-55. 16. Schulz R, Wuster M, Herz A. Centrally and peripherally mediated inhibition of intestinal motility by opioids. Naunyn-Schmiedebergs Arch Pharmacol 1979;308:255-60.

1480

SCHILLER ET AI..

17. Wuster M, Herz A. Opiate agonist action of antidiarrheal agents in vitro and in vivo-findings in support for selective action. Naunyn Schmiedebergs Arch Pharmacol 1978;301: 187-94. 18. Binder HJ, Laurenson J, Dobbins JW. Mechanism of loperamide's antidiarrheal action (abstr). Gastroenterology 1981; 80:1111. 19. Beubler'E, Lembeck F. Inhibition of stimulated fluid secretion in the rat small and large intestine by opiate agonists. Naunyn Schmiedebergs Arch Pharmacol 1979;306:113-18. 20. Sandhu BK, Tripp JH, Candy DCA, Harries JT. Loperamide: studies on its mechanism of action. Gut 1981;22:658-62. 21. Farack VM, Kautz U, Loeschke K. Loperamide reduces the intestinal secretion but not the mucosal cAMP accumulation induced by cholera-toxin. Naunyn Schmiedebergs Arch PharmacoI1981;317:178-9. 22. Hardcastle J, Hardcastle PT, Read NW, Redfern JS. The action of loperamide in inhibiting prostaglandin-induced intestinal secretion in the rat. Br J Pharmacol 1981;74:563-9. 23. Hughes S, Higgs NB, Turnberg LA. Loperamide has antisecretory activity in vivo in human jejunum. Gut 1983;24:A495. 24. Kachel GW, Ruppin H, Hagel J, Barina W, Meinhard M, Domschke W. Effect of loperamide on jejunal fluid transport and propulsion in man (abstr). Gastroenterology 1983;84: 1200. 25. Baker GF, Segal MB. The effect of loperamide on the ion fluxes across the isolated rabbit colon. Biochem Pharmacol 1981;30:3371-3.

GASTROENTEROLOGY Vol. 86, No.6

26. Ilundain A, Naftalin RJ. Opiates increase chloride permeability of the serosal border of rabbit ileum. J Physiol (Lon d) 1981;316:56P-57P. 27. Hughes S, Higgs NB, Turnberg LA. Anti-diarrhoeal activity of loperamide: studies of its influence on ion transport across rabbit ileal mucosa in vitro. Gut 1982;23:974-9. 28. Davis GR, Santa Ana CA, Morawski SG, Fordtran JS. Development of a lavage solution associated with minimal water and electrolyte absorption or secretion. Gastroenterology 1980; 78:991-5. 29. Davis GR, Santa Ana CA, Morawski SG, Fordtran JS. Effect of vasoactive intestinal polypeptide on active and passive transport in human jejunum. J Clin Invest 1981;67:1687-94. 30. Hyden S. A turbidometric method for the determination of higher polyethylene glycols in biological materials. Lantbrukshogsk Ann 1955;22:139-45. 31. Michiels M, Hendriks R, Heykants J. Radioimmunoassay of the antidiarrhoealloperamide. Life Sci 1977;21:451-60. 32. Colton T. Statistics in medicine. Boston: Little Brown & Co., 1974. 33. Binder HJ. Absorption and secretion of water and electrolytes by small and large intestine. In: Sleisengei' MH, Fordtran JS, eds. Gastrointestinal disease. Pathophysiology. Diagnosis. Management. Philadelphia: WB Saunders, 1983:811-29. 34. Verhaeren EHC, Dreesen MJ, Lemli JA. Influence of 1,8dihydroxyanthraquinone and loperamide on the paracellular permeability across colonic mucosa. J Ph arm Pharmacol 1981;33:526-8.