Studies of the Function of the Human Pylorus : and Its Role in The

+.1

Studúes OlTlæ Ftrnctíon OJTIrc Humanfolonts

And,Iß R.ole InTlæ Riegulø;tíon OÍ

Cústríß Drnptging

David R. Fone

Departments of Medicine and Gastroenterology, Royal Adelaide Hospital

University of Adelaide

August 1990 Table of Contents

TABLE OF CONTENTS .

SUMMARY vil

DECLARATION...... X

DED|CAT|ON...... xt

ACKNOWLEDGMENTS xil

CHAPTER 1 ANATOMY OF THE PYLORUS

1.1 INTRODUCTION.. 1

1.2 MUSCULAR ANATOMY 2

1.3 MUCOSAL ANATOMY 4

1.4 NEURALANATOMY 1.4.1 Extrinsic lnnervation of the Pylorus 5 1.4.2 lntrinsic lnnervation of the Pylorus 7

1.5 INTERSTITIAL CELLS OF CAJAL 8

1.6 CONCLUSTON 9

CHAPTER 2 MEASUREMENT OF PYLORIC MOTILITY

2.1 INTRODUCTION 10

2.2 METHODOLOG ICAL CHALLENGES 2.2.1 The Anatomical Mobility of the Pylorus . . . 10 2.2.2 The Narrowness of the Zone of Pyloric Contraction 12

2.3 METHODS USED TO MEASURE PYLORIC MOTILITY 2.3.1 lntraluminal Techniques 2.3.1.1 Balloon Measurements...... 12

t 2.3 1.2 lntraluminal Side-hole Manometry . . 13 2.3 1.9 The Sleeve Sensor 14 2.3 1.4 Endoscopy. 16 2.3 1.5 Measurements of Transpyloric Flow . 16 2.3 'I .6 lmpedance Electrodes 16 2.3.2 Extraluminal Techniques For Recording Pyl;'; l'¡"r¡iit¡l 2.3.2.'t Strain Gauges ...... 17 2.3.2.2 lnduction Coils . . 17 2.3.2.3 Electromyography 17 2.3.3 Non-lnvasive Approaches For Recording 2.3.3.1 Radiology :ï:: Y:1":'1 18 2.3.9.2 Ultrasonography . . 1B 2.3.3.3 Electrogastrography 19 2.3.4 ln Vitro Studies of Pyloric Muscle 19

2.4 CONCLUSTON. 19

CHAPTER 3 MEASUREMENT OF GASTR¡C EMPTYING

3.1 INTRODUCTION. . 20

3.2 INVASIVETECHNIQUES FOR MEASURING GASTRIC EMPryING IN HUMANS 3.2.1 lntubation/Aspiration of Gastric Contents. 20 aã NON-INVASIVE TECHNIQUES FOR MEASURING GASTRIC EMPryING 3.3.'l Radiology 21 3.3.2 Ultrasonography . . 22 3.3.3 Applied Potential Tomography and lmpedance Epigastrography 22 3.3.4 Absorption Kinetics of Orally Administered Drugs . 22 3.3.5 RadionuclideTechniques 23 3.3.5.1 Radionuclide Markers 23 3.3.5.2 OtherMethodological lssues 25 s.s.s.s Data Acquisition and Processing 27 3.3.5.4 Geometrical Errors of Radionuclide Methods 27 3.3.5.5 Limitations of Radionuclide Methods 28

3.4 CONCLUSTON 28

CHAPTER 4 COORDINATED PATTERNS OF PYLORIC MOTILITY

4.1 INTRODUCTION.. 29 4.1.1 Definition of Antropyloroduodenal Coordination 30

4.2 COORDINATED PYLORIC MOTILITY DURING FASTING. . . 30 4.2.1 Patterns of Propagated Motility During Fasting 30 4.2.2 Localized Pyloric Motility During Fasting . . . . 34

4.3 COORDINATED PYLORIC MOTILITY IN THE FED STATE . . . 35 4.3.1 Patterns of Propagated Motility in the Fed State 35 4.3.2 Localized Pyloric Motility in the Fed State . . . . 37

4,4 CONCLUSION 39

It CHAPTER 5 CONTROL OF PYLORIC MOTILITY s.1 INTRODUCTION. 40

5.2 ELECTROMECHANICAL PFìOPERTIES OF PYLORIC SÍUOOTH MUSCLE 5.2.1 Characteristics of the Electrical Control and Response Activity of Smoolh Muscle Cells at the Distal Pyloric Muscle Loop . . . . . 42 5.2.2 Transmission of Electrical Activity Across the Pylorus 44 5.2.3 Mechanisms for the Modulation of Pyloric Electrical Activity by Neurotransmitters and Gastrointestinal Hormones. 46 5.2.4 Conclusion 46

5.3 CENTRAL NERVOUS SYSTEM CONTROLOF PYLORIC T/OTOR FUNCTION 46 5.3.1 Evidence for Control of the Pylorus by the Vagus Nerve . . . 47 5.3.1.1 Excitatory lnnervation of the Pylorus by the Vagus . . . 47 5.3.1.2 lnhibitory lnnervation of the Pylorus by the Vagus . . . . 47 5.3.2 Sympathetic Control of Pyloric Motility 48 5.4 CONTROL OF THE PYLORUS BY THE SMALL INTESTINE 49 5.4.1 The Effect of lntraluminal Small lntestinal Stimuli on Pyloric Motility 49 5.4.1.1 lntraluminal Nutrients 49 5.4.1.2 Small lntestinal Acidification . . 50 5.4.2 Sile and Nature of Mucosal Receptors in the Small lntestine . . 50 5.4.3 Candidate Local Neural Mechanisms for Mediating the Small lntestinal Control of Pyloric Motilily 51 5.4.3.1 Neural Pathways Connecting the Duodenum to the Pylorus. 52 5.4.3.2 Pathways Connecting the Antrum to the Pylorus . 52 5.4.3.3 Conclusion 53

5.5 C,ONTROLOF THE PYLORUS BY GASTROINTESTINAL HORMONES 53 5.5.1 Cholecystokinin . . 53 5.5.2 Secretin 55 5.s.3 Peptide YY . . . .55 5.5.4 Olher Candidale Hormones . . . .56 5.5.5 Summary .56 s.6 CONCLUSTON 57

CHAPTER 6 THE MECHANICAL ROLE OF THE PYLORUS

6.1 INTRODUCTION. . 58

6.2 THE PYLORUS AS A GASTROINTESTINAL SPHINCTER 59

6.3 THE MECHANICAL EFFECT OF LOCALIZED PYLORIC CONTRACTION . . . . . 60 6.3.1 The Pylorus and Liquid Gastric Emptying 60 6.3.2 The Pylorus and Gastric Emplying of Digestible Solids 62 6.3.3 The Gastric Emptying of Non-Digestible Solids During Fasting 65 6.3.4 The Pylorus and Duodenogastric Reflux 65

6.4 CONCLUSION 65

itt CHAPTER 7 METHODS USED TO STUDY PYLORIC MOTILITY AND GASTRIC EMPTYING IN THIS THESIS

7.1 INTRODUCTION. 66

7.2 SELECTION OF SUBJECTS 66

7.3 i/IANOI/ETRIC TECH N IQIJ E 7.3.1 lntubationTechnique .67 7.3.2 Manometric Methods 7.3.2.1 Design of the Manometric Assembly 67 7.3.2.2 Recording of lntraluminal Pressures 69

7.3.3 Measurement of Transmucosal Potential Difference 7.s.s.'l Background 70 7.3.3.2 Technique 70

7.4 MEASUREMENT OF GASTRIC EMPTYING 72

7.5 DATA ANALYSIS 7.5.1 Analysis of Recordings . . . . 72 7.5.2 Analysis of Gastric Emptying Measurements 75 7.5.3 Statistical Tests . 75

CHAPTER 8 TECHNICAL ASPECTS INVOLVED IN INTERPRETING COMBINED SLEEVE/SIDE.HOLE PYLORIC MANOMETRY IN HUMANS

8.1 INTRODUCTION.. .. 76

8.2 A STUDY OF THE COMPARATIVE EFFECTS OF DUODENAL AND ILEAL INTUBATION ON PYLORIC, ANTRAL AND DUODENAL h/OTILITY AND THE GASTRIC EMPTYING OF A SOLID MEAL 8.2.1 Background 76 8.2.2 Materials and Melhods 8.2.2.1 Subjects 77 8.2.2.2 Experimental Protocol 77 8.2.2.9 Data Analysis . . . . 78 8.2.3 Results 78 8.2.3.1 GastricEmptying 78 8.2.3.2 Motility 81 8.2.3.3 Relationships Between Gastric Emptying, Motility and lntubation Time B2 8.2.4 Summary of Results 82

8.3 AN EVALUATION OF THE NORMAL PATTERNS OF HUMAN PYLORIC AND ANTRAL I/OTILITY WITH A NOVEL ANTRAL WALL MOTION DETECTOR 8.3.1 lntroduction 82 8.3.2 Materials and Methods 8.3.2.r subjects 83 8.3.2.2 Experimental Protocol .83

tv 8.3.2.3 Measurement of Motility . . 83 8.3.3 Data Analysis 8.3.3.1 Analysis of Records 84 8.3.3.2 Statistical Methods 87

8.3.4 Resulls 8.3.4.1 Subject Tolerance and Sensor Positioning 87 8.3.4.2 Antral Motility . . . . 87 8.3.4.3 lsolated Pyloric Pressure Waves 92 8.3.5 Summary of Results 96

8.4 DISCUSSION 96

8.s CONCLUSTON 102

CHAPTER 9 CONTROL OF PYLORIC MOTILITY BY STIMULATION OF SMALL INTEST¡NAL RECEPTORS

9.1 INTRODUCTION 103

9.2 A STUDY OF THE ROLE OF MUSCARINIC MECHANISMS IN THE PYLORIC i/lOTOR RESPONSE TO INTRADIJODENAL DEXTFìOSE IN HUMANS 9.2.1 Background 103 9.2.2 Methods 9.2.2.1 Subjects 104 9.2.2.2 Experimental Design . 104 9.2.2.3 Pressure Measurement 106 9.2.2.4 DalaAnalysis...... 106 9.2.2.s Statistical Analysis . 106 9.2.3 Results 106 9.2.3.1 Pyloric Motor Responses lo lntraduodenal Dextrose 107 9.2.3.2 Antroduodenal Motor Changes 112 9.2.4 Summary of Results 114

9.3 A STUDY OF THE EFFECT OF DISTAL SMALL INTESTINALTRIGLYCERIDE INFUSION ON PYLORIC, ANTRAL AND DUODENAL MOTILITY, GASTRIC EMPTYING AND THE INTRAGASTRIC DISTRIBUTION OF A SOLID MEAL 9.3.1 Background 114 9.3.2 Materials and Methods 9.3.2.1 Subjects 115 9.3.2.2 Experimental Protocol 115 9.3.2.3 Manometric Methods . 118 9.s.2.4 Data Analysis ...... 118 L9.2.5 Statistical Analysis . . 118 9.3.3 Results 9.3.3.1 Subject Tolerance and Assembly Position 119 9.s.s.2 Gastric Emplying Profile 119 9.3.3.3 Antropyloroduodenal Motor Response . . 124 9.3.3.4 Correlation Between Gastric Emptying and Motility . . . 124 9.3.4 Summary of Results '127

9.4 DISCUSSION 128 e.s coNcLUStoN 132

Y CHAPTER 10 A STUDY OF PYLORIC MOTILITY DURING THE DELAYED GASTRIC EMPTYING INDUCED BY COLD STRESS

10.1 |NTRODUCTION.. 134

10.2 i/ATERIALS AND METHODS '1O.2.1 Subjecls ... 134 10.2.2 Experimental Protocol ... 135 10.2.3 Manometric Methods .. . 136 10.2.4 Data and Statistical Analysis ... 136

10.3 RESULTS 10.3.1 Subject Tolerance of Cold Stress 136 10.3.2 Gastric Emptying 137 10.3.3 Pyloric MotilitY 141 10.3.4 Antral and Duodenal Motility 10.3.4.1 Antral Motility . . . . 141 10.3.4.2 Duodenal MotilitY 141 10.3.5 Relationship Between Motility and Gastric Emptying . '147

10.4 SUMMARY OF STUDY 148

10.5 DISCUSSION 149

10.6 CONCLUSION 150

CHAPTER 11 FUTURE MANOMETR¡C INVEST¡GAT¡ON OF NORMAL PYLORIC MOTOR FUNCTION IN HUMANS

11.1 INTRODUCTION 1s1

11.2 RECONíIúENDATIONS FOR IT/PROVED i/ETHODOLOGICAL APPROACHES FOR PYLORICMANOMETRY.... 151

11.3 INVESTIGATION OF THE COORDINATED PATTERNS AND PHYSIOLOGICAL CONTROL OF HUMAN PYLORIC MOTILITY 152

11.4 THE IMPORTANCE OF THE HUI/AN PYLORUS AND COORDINATED PYLORIC TTOTILITY IN THE REGULATION OF GASTRIC EMPTYING . . . . 154

11.s CONCLUSION . . 156

APPENDIX 157

BIBLIOGRAPHY 159

Yl Summary

This thesis presents studies performed in healthy human volunleers which examine aspects of the control and measurement of pyloric motor function believed to be relevant to the role(s) of the pylorus in the regulation of normal gastric emptying. There is limited and contradictory information about pyloric motor function, largely because of the considerable technical challenges associated with accurate measurement of pyloric motility. The recent development of the perfused sleeve sensor has enabled reliable, prolonged pyloric manometry in humans for the first time. lnsights gained wilh combination sleeve and multiple manometric side-hole studies indicate that lhe majority of measurement techniques used in the past have been unsatisfactory, because they did not account for the mobility of the gastroduodenal junction and the narrowness of the zone of pyloric contractions. Recent studies with sleeve/side-hole manometric assemblies have documented more precisely the patterns of pyloric, antral and duodenal contractions in health, and have provided the most relevant background information to the studies performed in the presont project with the same melhodology. A number of the manometric studies described in this thesis were also performed with simultaneous radionuclide measurements of gastric emptying and the intragastric distribution of a solid meal.

It is contentious whether isolated pyloric pressure waves, which have been recorded with sleeve/side-hole manometric assemblies and occur in the absence of detectable changes in intraluminal pressure in the antrum and duodenum, are due to truly isolated contractions of the pylorus or, alternatively, are associated with undetected low-amplitude antral contractions. This issue was investigated with a novel intraluminal sensor, which included an elliptical wire spring designed to detect low amplitude movements of the distal anlral wall. This wall movement detector was incorporated into a sleeve/side-hole manometric catheter, and was found to be more sensitive than intraluminal manometry for detecting antral contractions. Use of this device led to the confirmation that manometrically defined isolated pyloric pressure waves are nearly always due lo localized pyloric contraction, and also gave new insights into better approaches for the interpretation of intraluminal sleeve/side-hole manomelric recordings. lnlormation gained

vit lrom the transducer tracings suggêsted improvements in the spacings between side-holes on sleeve manomelric assemblies which would enable more accurate evaluation of the patterns of antropyloroduodenal motility.

There are conflicting reports in the literature about the effects of gastrointestinal intubation on gastric emptying and a lack of information about the effects of intubation on gastric, pyloric or duodenal motility. Accordingly, studies were performed which measured pyloric, antral and duodenal motility and gastric emptying of a digestible solid meal after naso-duodenal and naso- ileal intubation. Gastric emptying was measured in a further control group of subjects who were not intubated. There was no significant effect ol duodenal or ileal intubat¡on on pyloric motility. lleal, but not duodenal, intubation was associated with a significant decrease in the number of antral pressure waves and in the rate of gastric emptying. These results suggest that, while there is no significant effect of duodenal intubation on gastric emptying, antral or pyloric motility, the effects of intubation of the distal small intestine on both gastroduodenal motility and gastric emptying must be considered and controlled for.

The presence of nutrients within the small intestine is known to slow gastric emptying. Stimulation of phasic and tonic pyloric contraction and inhibition of antral contractions are possible motor mechanisms responsible for this effect, since intraduodenal nutrient infusions, amongst them dextrose, produce this motor pattern. The role of muscarinic mechanisms in the stimulation of the pylorus by intraduodenal dextrose was assessed by examining the effect of intravenous atropine on this response. The phasic and the tonic pyloric motor resPonses observed after intraduodenal dextrose were significantly reduced by atropine. Recognition of the muscarinic dependency of the pyloric motor response to intraduodenal dextrose provides new insights into the potential neural pathways and candidate hormonal factors which control pyloric motility.

The infusion of lipid into the distal small intestine has also been shown to delay gastric emptying. The motor mechanisms responsible for this effect were similarly investigated. lnfusion of a triglyceride emulsion into the terminal ileum was confirmed to slow gastric emPtying and, in association with this effect, there was a significant increase in the number of isolated pyloric pressure waves and reductions in antral and duodenal motility. Retrograde movement of part of the meal from the distal to the proximal stomach was demonstrated in some subjects in association with the slowing of gastric emptying. These results indicate that the pylorus is responsive to stimulation by luminal nutrient receptors beyond the proximal small bowel, and supports the view that the pylorus contributes to the regulation of the rate of gastric emptying.

vltt ln a further series of experiments, cold pain, an acute cenlrally acting stimulus, was shown to result in changes in pyloric, antral and duodenal motility, gastric emptying and intragastric distribution of a solid meal similar to those observed after ileal triglyceride infusion. These observations demonstrate lhat csntral nervous system signals are also important in the control of pyloric motility in humans.

Methodological insights gained from experiments described in this thesis enable improved interpretation of intraluminal sleeve/side-hole pyloric manometry in humans. ln addition, new information has been provided about the intestinal and central mechanisms controlling the pylorus, with the results suggesting an important role for the pylorus in the regulation of gastric emptying.

lx rl

Declaratlon . . .

The experimental work presented in this thesis is entirely original and was performed by the author in 1987 and 1989. The thesis contains no material which has been accepted for the award of any other degree or diploma in any University. The results have not been previously published except by the author as scientific papers directly resulting from this work. Figures used in the literature review were included with the kind permission of relevant authors and publishers, as thesis may be photocopied, and the thesis |f indicated in the text. Material contained within this ,lj may be made available for loan at the discretion of the University.

x Dedication . . .

To Jennifer, Alexandra and Timothy, without whose encouragement, lolerance and good faith this work would not have been completed.

',| ,l!

d ! Acknowledgments

The work reported in this thesis is submitted for the degree of Doctor of Medicine to the University of Adelaide. All studies were performed within the Departments of Gastroenterology, Medicine and Nuclear Medicine al the Royal Adelaide Hospital during 1987 and 1988. Throughout this time, and subsequently during the analysis and preparation of results for publication and the writing of this lhesis, I received invaluable support from Dr. Michael Horowitz, from the Department of Medicine, and from Professor John Dent, Director of Gastroenlerology. Their advice was at all limes friendly and freely given, their criticisms conslructive, and their reasoning incisive and well focussed. They helped to provide the stimulating environment necessary for productive scientific work, and when required, their encouragement bolstered my enthusiasm. On numerous occcasions I enjoyed discussions on theoretical aspects of manometry .'+ 'rJ with Professor Dent who, as the original designer of the sleeve sensor, spoke with considerable l authority. Similarly, Dr. Horowitz was able to help with problems relating to the radionuclide measurement of gastric emptying. I owe both a great debt of gratitude.

I would also like to thank Dr. Richard Heddle and Dr. Robert Fraser. Dr. Heddle was involved in the initial development and application of the recording equipment used in this thesis, and first introduced me to the practical complexities of pyloric manometry. His thorough attention lo detail and complete objectivity were particularly instructive. His advice and later his friendship were greatly appreciated. Dr. Fraser was also a colleague in the motility laboratory, and we were involved in several projects together. He was always willing to discuss any concerns arising from my work and provided scientific and moral support. Special mention should be made of Professors Louis Akkermans of The Netherlands and Nick Read from England, who provided invaluable insights and specific expertise for several studies. Advice about the design, analysis and reporting of experiments was also gratefully received from Dr. Richard Holloway and Dr. lan Cook. My gratitude is also extended to Dr. Peter van Rijk, Director of the Department of Nuclear I Medicine at the Academic Hospital Utrecht, for making available the excellent computing facilities of his department during the writing and final compilation of this thesis.

dt I Facilities in the Department of Nuclear Medicine at the Royal Adelaide Hospital were required for lwo sels of experimenls. Through the efforts and cooperation of Dr. Barry Chatterton, lhe Director of this department, a gamma camera and computer were always made available, despite considerable clinical demands. On several occasions, advice was provided by Mr. Peter Collins who, together with Drs. Michael Horowitz and Barry Chatterton, had developed the method used in lhis thesis for radionuclide measurement of the distribulion of food within the stomach. Mosl important was the assistance of Mrs. Anne Maddox, who provided expert technical support during the performance and analysis of gastric emptying sludies. Anne's ability, enthusiasm, professional attitude and unfailing good humour ensured the success of these studies.

Special thanks also go to several people who provided equipment for the measurement of motility. I am especially grateful again to Professor Dent for the manufaclure ol all sleeve/side-hole manometric assemblies. The antral wall motion transducer used in one study was developed at the Rotterdam Academic Hospital by Professor Louis Akkermans and Evert van der Schee. Equipment used for intraluminal manometry was meticulously maintained by the laboralory manager Mr. Marcus Tippett. He and Miss Petra Kocyan provided technical support for manometric studies. Together wilh Miss Sue Graham, Voula Panagopolous and Maria Simula, lhey created an enjoyable and relaxed atmosphere in the laboratory.

possible the .l Completion of the studies reported in this thesis would not have been without 'it'.[l assistance of lhose mentioned above. Their friendship will always be remembered. I

f I

*lll ! CHAPTER 1

Anatomy of the PYlorus

1.1 INTRODUCTION The morphology of the pylorus in a number of species, including man, has been well reviewed by Torgersen (1542) and Sehulze-Delrieu ei al (i984). The anatorny of the pylorus, and its relationship to the adjacent anlrum and duodenum, determines to a large extent the function of the pylorus (Chapter 6) and the patterns of pyloric contractions (Chapter 4). Many of the technical difficulties associated with accurate recording of pyloric motility (Chapter 2) are explicable on the basis of the special arrangements of the pyloric muscle loops. Furthermore, studies of the microscopic sfructure and relationships of smooth muscle, nerve and other cells contained wilhin the pylorus provide insights into the mechanisms mediating the coordination of antropyloroduodenal motility (Chapter 5). This chapter reviews the literature on the macroscopic and microscopic anatomy of the pylorus, as well as its intrinsic and extrinsic neural innerualion' Special attention is paid to those features which are important in understanding the above issues.

The term *pylorus" has been used inconsistently in the literature to refer to different analomical structures in the region of the gastroduodenal junction. The definitions proposed by Schulze- Delrieu et al (1984), based on their studies of the human pylorus and structural studies in a number of species by Torgersen (1942), will be used in the first section of this thesis (Figure 1.1). Although in a strict anatomical sense the term "pylorus" is usually applied to the two pyloric muscle loops and the area between, il will be used subsequently in this thesis to indicate the distal pyloric muscle loop. This slruclure, because of its site, dimensions and specialized innervation, is likely fo be responsible for lhe manometric phenomena attributed to the 'pylorus' i (Chapters 2 and 4). t 2

1.2 MUSCULAR ANATOMY The pylorus consists of a well-defined thickening of smooth muscle at the gastroduodenal junction (Figure 1.1). Since the oblique layer of the gastric musculature is poorly represented in lhe immediately orad distal antrum, the pylorus is formed predominantly by a specialized arrangement of the outer longitudinal and the inner circular smooth muscle coats (Williams and Warwick 1g8O). Macroscopically, at the level of the human gastroduodenal junction there is a ring of circular smooth muscle, which is up to 1.5 cm wide (Horton 1928; Torgersen 1942). This ring is continuous with the inner circular muscle layer of the gaslric antrum (Williams and Warwick 1980) and is known as the distal pyloric muscle loop. The greatesl density of circular smooth muscle fibres is found on the lesser curve aspect, where the distal pyloric loop is about 5 to 15 mm wide (Torgersen 1942) and 3.7 to 14.0 mm thick (Schulze-Delrieu and Shirazi 1983). From the lesser curve the circular fibres fan out towards the greater curve, so that at lhe greater curve the pylorus is 20 to 35 mm wide (Schulze-Delrieu et al 1984; Torgersen 1942). At the proximal edge of lhis radiation, the circular fibres form a second thickening, the proximal pyloric loop. On the lesser curue aspect, the proximal and distal pyloric loops are confluent and, together with the abundant fat and connective tissue of the submucosa, form a prominent "knot" known as the "pyloric torus". Between the two pyloric loops on the greater curve aspect is a relatively thin area of circular muscle which connects the two rings, so that they are anatomically dependent, and into which the pyloric torus fits snugly when the pylorus contracts (Schulze-Delrieu et al 1984) (Figure 1.2). The segment of gastroduodenal junction between the two loops is called the'pyloric segmentn (Torgersen 1942; Schulze-Delrieu and Shirazi 1983); the lumen of the contracted pyloric segment is known as the 'pyloric canal".

Longitudinal smooth muscle also contributes to the pylorus. There are two sets of outer longitudinal fibres in the stomach wall, one which runs from the gastro-oesophageal junction over the cardia, and one which extends from the body of the stomach to the gastroduodenal junction. ln the proximal antrum the longitudinal layer becomes thicker and runs distally over the proximal and distal pyloric loops, forming a band which is thicker on thê greater curvaturê aspect and relatively thin anteriorly and posteriorly. The relationship between the longitudinal and circular smooth muscle of the distal antrum and that of the pylorus is likely to be partly responsible for the ability of the pylorus to frequently contract in sequence, or simultaneously, with the terminal antrum (Chapter 4).

The majority of the longitudinal fibres dip into the connective tissue and interlace with the circular muscle fibres of the distal pyloric loop (Horton 1928; Torgersen 1942). From multiple longitudinal sections of the human gastroduodenal junction it has been estimated that only 21- 24/o o1 these longitudinal fibres are continuous with the longitudinal muscle of the duodenum (Horton 1928). Continuity of longitudinal smooth muscle is most apparent macroscopically on the 3

A) B) Pyloric Segnent Atrtrur. Mu¡ob Duode¡un MutcL A.¡trum

Termin¿l ¡ntrum Prorimal Pyloric Muscle Loop Torug Torus Pyloric D¡strl Pyloric Orifice Duodenun Muscle Loop Proximel Pylcic Dist¡l Pyloric Murcle Loop Müscle LooP

Flgure 1.1 junction Muscular anatomy of the human pylorus. On the left (A) is a sketch of the gastroduodenal opened along the greater curvature. The orad end of the pylorus is marked by fhe Proximal pylor'lc muscle loop, (B) whicÉ joins the distal pyloric muscle loop on the lesser curvature to form the pyloric torus. On the right is a longitudinal section through the pylorus. (Adapted from Schulze-Delrieu and Shirazi 1983, and Schulze- Delrieu et al 1984, with permission.)

Py lor¡c P ylor¡c oma Torus Torus Stomech

PPL B OB D L PPL OPL

Rels r ed Coñlraclcd

Figure 1.2 Schematic representation of the relaxed (left) and contracted (right) human pylorus. DB = duodenal bulb; DPL pylorus of the lwo musde = distal pylo¡c loop; PPL = proximal pyloric loop. ln the contracted the shortening loops leads to the partial formation of the pyloric canal. Further conlraction would oblitsrate the canal. (From Schulze-Delrieu et al 1984, with permission') 4

lesser curve (Torgersen 1942). The termination of much of the longitudinal muscle layer at the pylorus, the relative thinness of the circular muscle coat and the absence of the muscularis mucosae (Torgersen '1942), contribute to the sparsity of muscle in the proximal duodenal bulb (Schulze-Delrieu 1984), thus called the 'hypomuscular segment" (Code 1970)'

A number of histological studies have reporled lhat, in contrast to the partial continuity of the longitudinal smooth muscle, the circular muscle of the pylorus is separated entirely from thal of the duodenal bulb by a connective tissue septum immediately distal to the distal pyloric loop (Horton 1928; Code 1970; Williams and Warwick 1980). Torgersen (19421, however, disputed the existence of such a separation, since he interpreted his dissections as showing continuity of the circular smooth muscle layer between the pylorus and the duodenum on the lesser curve aspect. This discrepancy is likely to have arisen because of difficulties in the macroscopic (versus microscopic) assessment of muscle fibre connections. The relative discontinuity between pyloric and duodenal smooth muscle was thought by Horton (1928 & 1931) to be responsible for 'pyloric block' of propagating antral contractions, preventing their continuation into the duodenum (Cannon 1898). The fact that many contractions do, however, propagate across the pylorus (Chapter 4) may be explained by the transmission of antropyloric electrical activity into the duodenum via longitudinal smooth muscle bundles (Chapter 5), although transmission could occur also via the myenteric plexus, or possibly a network of interstitial cells of Cajal (see Seclions 1.4.2 and 1.5).

There is a paucity of information about the microsopic anatomy of pyloric smooth muscle cells. Recently, however, Daniel and Allescher (1990) reported that the muscle cells of the dislal pyloric loop in dogs have several important differences from those found elsewhere in the gastrointestinal tract. For example, pyloric smooth muscle cells have a higher density of nerves (3-5 fold) in close proximily, compared lo the canine corpus (see below). Furthermore, while pyloric muscle cells have a moderately high density of gap junctions and other junctions, such as close appositions and intermediate contacts, there are probably fewer gap junctions than in the antrum or fundus (Oki and Daniel 1974; Daniel and Allescher 1990). ln the outer portion of the distal pyloric loop closest to the duodenum, the gap junctions apPsar smaller. The functional significance of these observations remains unclear.

1.3 MUCOSAL ANATOMY The mucous membrane is firmly attached to the distal pyloric muscle loop by fibrous lissue which is continuous with lhe connective tissue septum between the circular smooth muscle of the pylorus and the duodenum (Williams and Warwick 1980), and by muscle bands from the longitudinal layer which invade and sometimes penetrate the circular muscle ring to reach the mucosa (Torgersen 1942; Schulze-Delrieu et al 1984). At the level of the distal pyloric muscle 5

loop, the muscularis propria and the muscularis mucosae fuse, resulting in close attachment of the mucosa to the submucosa and muscle layers. This attachment has been demonstrated in vitro by the limitation of the spread of lndian ink, injected under pressure into the antral submucosa, at the level of the distal pyloric loop (Horton 1931; Schulze-Delrieu and Shirazi 1983).

The mucosa of the duodenum is also firmly attached to the underlying muscle. The glands of Brunner, which penetrate deep into the submucosa (Alvarez 1940), may contribute lo this anchorage. ln the distal antrum, however, the mucosa is loosely attached, forming longitudinal folds which project into the adjacent orad antrum and aborad duodenum when the pylorus contracts. Radiological and in vitro studies suggest that this "redundant" mucosa may act as a watert¡ght plug (Williams 1962; Biancani et al 1980). Tall columnar mucus-secreting cells of the pyloric canal give way to the columnar duodenal epithelium at, or near the level of the connective tissue septum. The different properties of these cells results in an abrupt change in electrical potential difference at this level (Andersson and Grossman 1965; Fisher and Cohen 1973).

1.4 NEURAL ANATOMY

1.4.1 Extrlns¡c Innervatlon of the Pylorus There is limited information on the extrinsic nerve supply to the pylorus, particularly with regard to central neural projeclions. Dissections indicate that lhe pylorus is innervaled by lhe vagus nerve and by branches from the upper abdominal sympathetic plexuses.

Vagal molor and sensory fibres reach the pylorus via the two major groups of branches formed after division of the anterior vagal trunk below the diaphragm (Skandalakis et al 1986; Gray 1989). The majority of vagal fibres which innervate the pylorus travel via the left branch, or anterior gastric nerve, which supplies the lesser curye and antero-superior portion of the stomach including the pylorus distally, often via a major branch known as the nerve of Latarget. Lower rami of the right, or hepatic, group of branches arising from the anterior vagal trunk supply the pylorus and proximal duodenum. The posterior vagal trunk supplies the postero- inferior portion of the stomach, usually with the exception of the pyloric segment.

The pylorus receives sympathetic fibres from the coeliac plexus, mainly through the plexuses around the gastric and gaslroepiploic arteries. Branches from the hepatic plexus give rise to branches which accompany the right gastric artery to the pylorus, the gastroduodenal artery to the pylorus and duodenum, and the right gastroepiploic afiery to the greater curve (Williams and Warwick 1980). The left gastric plexus, which accompanies the left gastric artery along the lesser curve and joins gastric branches of the vagus, contains excitatory fibres to the pylorus and inhibitory fibres to the antrum. Some sympathetic fibres may reach the pylorus by 6

traversing the vagi, which acquire a number of adrenergic sympathetic fibres from the superior cervical and probably the stellate ganglia as they descend through the neck (Leidberg et al 1973; Agostoni et al 1975). Furthermore, lhe vagi may acquire sympalhetic fibres which innervate the pylorus from branches of thoracic sympathetic ganglia, which send branches to form oesophageal plexuses with the vagi in the thorax, before the vagi divide into the anterior and posterior vagal trunks (Williams and Warwick 1980; Skandalakis et al 1986). Labelling of neurons with horseradish peroxidase has also demonstrated sympathetic innervation of the pylorus by the coeliac/superior messnteric ganglion, the superior cervical ganglion and the stellate ganglion (Elfvin and Lindh 1982).

The most valuable information about the central connections of fibres which innervate the pylorus comes from studies in animals using retrograde tracing lechniques (Yammamoto 1977; Elfuin and Lindh 1982; Takayama 1982; Pagani et al 1985). Even so, in these studies lhere has been inadequate anatomical identification of the pylorus as the only structure innervated by lhe labelled neurones. For example, it has not been clearly stated whether lracer has been injecled specifically into the narrow distal pyloric muscle loop. ln addition, the more widespread central connections identified in some studies may have resulted from diffusion of lracer from the site of injection to adjacent areas or overlying antral muscle.

Tracing with horseradish peroxidase in guinea pigs (Elfvin and Lindh 1982) and fast blue in cats (Pagani et al 1985) has demonstrated that the extrinsic motor supply to lhe pylorus consists of preganglionic parasympathetic vagal fibres, arising within the dorsal motor nucleus of the vagus in the medullary brainstem. Although Elfvin and Lindh (1982) identified connections throughout the length of the rostrocaudal portion of this nucleus, Pagani et al (1985) found labelled cell bodies to be localized in a small area approximately 0.56 to 1.56 mm rostral to the obex. Electrical stimulation of the specific area just rostral to the obex in cats was shown to induce large amplitude contractions of the pylorus (Pagani et al 1985). This response was eliminated by ipsilateral vagotomy. A very localized distribution of cell bodies projecting to the pylorus is supported by the findings of retrograde tracing studies in the rat (Takayama 198.l).

Other studies in the cat suggest that some motor fibres may also project to the pylorus from the medial solitary nucleus (Yammamoto 19771, but this was not confirmed either by tracing or electrical stimulation in the cat (Pagani et al 1985), guinea pig or rat. This finding is likely to be due to imprecise labelling of neurones (see above), but it would also be consistent with some central neurons innervating other parts of the stomach as well as the pylorus, via collaterals (Elfuin and Lindh 1982). The frequent participation of the pylorus in coordinated contractions that sweep across the antrum, pylorus and duodenum (Chapter 4) would be compatible with some sharing of neural input between these adjacent areas. However, although the neurochemistry of 7

cêntral neurons controlling the pylorus has not been elucidated, the considerable evidence that the pylorus has a specialized neuropeptide complement (Chapter 5) and is capable of localized patterns of contraction (Chapter 4), suggests thal substantial overlap in innervation is unlikely.

The vagus was originally thought to contain only motor fibres, but it is now known that about 90% of abdominal vagal fibres are unmyelinated afferents (Agostoni ot al 1975). Horseradish peroxidase studies indicate that these sensory fibres travel from the pylorus along several routes (Elfvin and Lindh 1982). Vagal afferents arise from large numbers of mainly unipolar neurons in the nodose (inferior) ganglia and a small number of neurons from the jugular (superior) ganglia at the base of the cranium. Sensory neurons are also found in the dorsal root ganglia of both sides, particularly at levels T5 to T10, but also at levels T1 to T2.

1.4.2 lntrlnsic Innervatlon of the Pylorus Histological and immunocytochemical studies have demonstrated important dilferences between lhe intrinsic neural plexuses at the pylorus and those elsewhere in the gut wall. Auerbach's plexus is a continuous ganglionated nerve plexus supplying the smooth muscle fibres of the outer longitudinal and inner circular muscle coats of the gastrointestinal tract (lrwin 1931)" Auerbach,s plexus lies between these lwo smooth muscle layers, sending branches to both. There are neural connections between Auerbach's plexus and Meissner's plexus, the ganglionated nerue plexus within the submucosa (Gabella 1987). Both plexuses are continuous between the antrum and duodenum, buf display considerable specialization at the pylorus' At this level, the number of fibres, the number and size of the ganglia and the number of cells in the ganglia of Auerbach's plexus ars increased. There are an eslimatsd 2O,OOO ganglion cells/cm2 at the pylorus compared to 3,500 cells/cmz at lhe cardia. Furthermore, in contrast to the fibres of the antral plexus, the interganglionic fibres within the pylorus cross at numerous points. Ganglia connected to the myenteric plexus have been observed deep within the circular muscle overlying and in the pyloric sphincter (Daniel, Costa and Furness, personal communication). Ganglia in the proximal duodenum are similar in morphology to those in the pylorus, except that they contain fewer cells.

Gabella (1972) reported that "near the pylorusn, vagal fibres anaslomose with only 2 or 3 ganglia within Auerbach's plexus, compared to the less discriminating situation in the stomach, where each vagal fibre may anastomose with 20 to 30 ganglia. Unfortunately, the precise site of origin of the tissue studied is not defined. Anastomoses may also occur between fibres from the vagus and those from the coeliac plexus. ln some species, including cals and dogs, small blood vessels lie near the ganglia, around which they may lorm capillary networks. Neuronal surfaces are direcfly exposed to connective tissue and to exlracellular spaces. Contraction or relaxation of muscle alters the shape and thickness of the ganglia, and Gabella has speculated that the shape and, possibly as a consequence, the function of ganglionic neural and glial cells may be similarly I

affêcted by mechanical forces (Gabella 1972'). Although unproven, such a process could add another interesting dimension to the variables which modulate pyloric motor function.

Detailed histological study of the canine pylorus has also shown a much higher density of neural varicosities compared with the antrum (Daniel and Allescher 1990). Many of these varicosities conlain a mixture of synaptic vesicles (Cai and Gabella 1984; Daniel and Allescher 1990). However, when compared with other tissues such as the lower oesophageal sphincter or gastric smooth muscle, a much higher proportion of pyloric neural varicosities contain a predominance of large, rather than small, granular vesicles (Oki and Daniel 1974; Daniel and Posey-Daniel 1984a; Daniel et al 19g9; Daniel and Allescher 1990). lmmunoreactiv¡ty for substance P, vasoactive intestinal polypeptide and probably other peptides including enkephalins, cholecystokinin and neuropepfide Y, has been demonstrated in the large granular vesicles of pyloric nerves (Daniel and Allescher 19g0). The increased density of these struclures at the pylorus strongly suggests that pyloric smooth muscle has a specialized and major input from peplidergic nerves' lnterestingly, almost none of the nerve profiles described above were in very close relation to smooth muscle cells, and very few closely innervated interstitial cells of Cajal (see below). possibly, these nerves may exert their influence by releasing mediators in the general environment of smooth muscle cells, interstitial cells of Cajal and other neural structures, rather than by direct synaptic contact.

1.5 INTERSTITIAL CELLS OF CAJAL pacemaker activify in the gastrointestinal tract was considered for many years to be myogen¡c in origin, since it occurs without neural inputs and persists in the presence of tetrodoloxin (Daniel and Sarna 1g78). Recently, it has been proposed that pacemaker activity for circular smooth muscle of the gut may instead be generated by interstitial cells of Cajal.

First clearly identified at an ultrastructural level by lmaizumi and Hama (1969), interstitial cells of Cajal are divided into several classes on the basis of their morphology and location. One type' for example, is located primarily within circular smooth muscle, and another is associated with ths myenteric plexus (see Thuneberg 1989 for a comprehensive review). Each of these two types of intersitial cell of Cajal forms a network with other similar inlerstitial cells and with smooth muscle cells, and both networks are closely associated with neural structures. lt was this arrangement which first prompted the suggestion that interstitial cells of Cajal may be involved in the origin, neurotransmission, or neuromodulation of gut pacemaker activity (lmaizumi and Hama 1969; Thuneberg 1982; Daniel and Posey-Daniel 1984a; Hara 1986).

The distribution of interstitial cells of Cajal throughout the gastrointestinal tract is not uniform, but thers have been no specific studies of their numbers, slructure or function in areas which I

have special properties, such as high slow wave frequencies. lnformation aboul interstitial cells of Cajal at the pylorus is primarily derived from a study in dogs by Daniel and Allescher (1990). There is no information about pyloric interstitial cells of Cajal in humans.

Electrophysiological studies have shown that the electrical control activity (Chapter 5) of the gut originates at or near the inner border of the circular smooth muscle layer, an area rich in interstitial cells of Cajal (Kobayashi et al 1966; Hara 1986). ln vitro observations with microelectrodes recently confirmed for the first time thal interstitial cells of Cajal are able to generate electrical activity (Barajas-Lopez et al 1989), and that generation of plateau potentials requires the coupling of interstitial cells of Cajal to circular smoolh muscle cells and/or one another. ln addition, interstitial cells of Cajal are the proposed site of release of nonadrenergic, noncholinergic inhibitory mediator(s) (Thuneberg 1989; Daniel and Allescher 1990) and are often closely innervated by nerves containing vasoactive intestinal polypeptide, a peptide present within the distal pyloric smooth muscle and a candidate hormone for the control of pyloric motor function (Chapter 5). These data stongly suggest that interstitial cells of Cajal, coupled to circular smooth muscle cells, are required for pacemaker activity (Thuneberg 1982 & 1989).

At the canine pylorus, interstitial cells of Cajal may be distinguished from those elsewhere on a number of grounds (Daniel and Allescher 1990). Firstly, they appear relatively sparsely scattered and do not usually lie close (within 30 nm) to nerves. Secondly, most are of the circular smooth muscle type, although a few are fibroblast-like, similar to those found elsewhere associated with the myenteric plexus. Thirdly, interstitial cells of Cajal at the pylorus form only occasional gap junctions with themselves, smooth muscle and nerves and, compared with those in the antrum or corpus, between two and eight times as many contain large nuclei. The existence of a network of interstitial cells of Cajal at the pylorus remains speculative, but such a network could facilitate transmission of electrical activity across the pylorus, thereby enabling the coordination of pyloric, antral and duodenal contractions (Chapters 4 and 5).

1.6 CONCLUSION The structure and innervation of the pylorus provide a logical basis for understanding the complex patterns of pyloric contractions, and their relationship to those of the antrum and duodenum. The smooth muscle of the pylorus is continuous with that of the antrum and connected wilh the duodenum by longitudinal smooth muscle bundles. There is conclusive evidence lhat specialization of smooth muscle and nerues occurs at the level of the distal pyloric muscle loop. Less well understood are the morphology and function of interstitial cells of Cajal at this level, although these cells have properties which may be critical to the function of the pylorus. The central connections of vagal efferents to the pylorus have been fairly well identified, but the sensory innervation of the distal pyloric muscle loop remains incompletely known. 10

CHAPTER 2

Measurement of Pyloric MotilitY

2.1 INTRODUCTION

Accurate, prolonged recording from the human pylorus Presents a considerable technical challenge. Recent detailed studies of the topography of isolated phasic and tonic pyloric contractions indicate that pyloric contractions produce a very narrow zone of increased pressure at the gastroduodenal junction (Heddle et al 'f 988b). Techniques for measuring pyloric motility must have the capability of recording from this specific, narrow zone. Furthermore, intraluminal sensors must allow for axial movsment of the mobile gastroduodenal junction over the sensor' These requirements were not satisfied by the methodological approaches of many early studies of pyloric motility, in which the technical demands of pyloric manometry were seriously underestimated. This deficiency has resulted in persistent coniroversy, not only about the patterns of motility displayed by the pylorus, but also about the mechanical role of the pylorus in the physiological control of gastric emptying.

This chapter will describe the methodological requirements for successful studies of pyloric motility, and the techniques used in previous studies (Table 2.1). Emphasis will be placed on those techniques suitable for intraluminal measurements of movement and pressure at the pylorus in

humans.

2.2 METHODOLOGICAL CHALLENGES 2.2.1 The Anatomical Mobility of the Pylorus The pylorus is an anatomically mobile structure although, unlike the upper and lower oesophageal sphinclers, there has been no formal evaluation of the degree of mobility which can occur. Nevertheless, there is considerable evidence that anatomical mobility of the pylorus is sufficient 11

1. Measurement of Electrical Activity ' electroPhYsiologY . electrogastrograPhY

2. Measurement of Smooth Muscle Contraction - in vitro (direct) ' isolated smooth muscle striPs - in vivo (indirect) ' strain gauges ' induction coils 3. lmaging Techniques . radiologY . ultrasonograPhY . endoscoPY

4. lntraluminal ManometrY ' Perfused side-holes . sleeve sensor ' non-perfused air- or water-filled channels . balloons

5. lndirect methods . measurement of transpyloric flow . imPedance electrodes

Tabþ 2.1 Techniques for measurement of pyloric motility

to present a significant challenge for the successful continuous recording of pyloric motility with intraluminal sensors. ln the experience of the author, movements of the pylorus of up to several centimetres relative to a sleeve sensor (Section 2.3.1.3) may be observed frequently during prolonged recordings. Spontaneous sliding of the narrow zone of pyloric conlraction over an intraluminal assembly was convincingly demonstrated by Heddle et al (1988b). Johnson (1981) found it impossible to keep a single, small intraluminal transducer wilhin the pylorus, despite the aid of fluoroscopy. ln cats, Reynolds et al (1984) found that even catheters positioned retrogradely across the gastroduodenal junction from the duodenum required pinning because of movement of the pylorus relalive 1o the sensor'

Successful measurements of pyloric molility can be achieved in animals by surgical fixation of the sensor at the level of the distal pyloric muscle loop (Behar et al 1979; Biancani et al 1981 ; Reynolds et al 1984 & 1985; Bertiger et al 1987; Ehrlein 1988). These methods are clearly inappropriate for human studies. Despite ingenious attempts (Aste et al 1979; Pandolfo et al 197g), no suitable technique of anchoring a sensor at the human pylorus is available. lf anchorage 12

at the pylorus is not possible, then the recording assembly must be able to tolerate a reasonable degree of axial movement without losing position.

position can A reliable method of determining sensor position is also required, so that any loss of be rapidly identified and corrected. Because significant movement of the pylorus may occur frequently and sometimes unprediclably, determining correct sensor position only at the al beginning of an experiment, or intermittently during recording (Aste et al 1979; Pandolfo et essential' 1979; Dooley et al 1985) is inadequate. Continuous feedback of assembly position is Continuous fluoroscopy cannot be used in humans because of the amounl of radiation exposure involved, however continuous measurements of transmucosal potential difference are feasible pull- and safe. Single-point measurements may be satisfactory for locating the pylorus during with rhrough recordings (Andersson and Grossman 1965), and have also been used in conjunction mucosal suct¡on to anchor the sensor (Pandolfo et al 1979). For prolonged recordings, dual-point proximal msasurements of transmucosal potential difference from the distal antrum and duodenum on either side of fhe manometric assembly are preferable, and have been used successfully in an number of studies in humans (Fisher and Cohen 1973; White et al 1981 ; Kaye et al 1976; Tougas et al 1987; Heddle et al 1988a-c; Houghton et al 1988aab)'

2.2.2 The Narrowness of the Zone of Pyloric Contractlon pyloric The second major challenge of recording pyloric motility is the narrowness ol the zone of (1988b)' This contraction, which was not fully appreciated until the recent study by Heddle et al detailed study of the topography of human pyloric contraction, with a manometric assembly active incorporating 13 side-holes at 3 mm intervals, demonstrated that the phasic and tonically (66%) phasic zone at the pylorus is almost always less than 9 mm wide, and for the majority of loop contractions is less than 6mm wide. The thickness and length of the distal pyloric muscle suggssts that it is fhis specific slruclure, rather than the broader structure containing the two loops and pyloric canal (Chapter 1), which is primarily responsible for generating such a narrow zone of increased intraluminal pressure. Manometric sensors for use in humans must be designed to record reliably from this narrow zone. The requirements necessary for accurate localization of the recording assembly within a mobile structure (Section 2.2.11 are also essential to meet this second challenge of recording from a very narrow zone'

2.3 METHODS USED TO MEASURE PYLORIC MOTILITY 2.3.1 Intralumlnal Techniques 2.3.1.1 Balloon Measurements lntraluminal balloons have been used in many sludies of pyloric, as well as antral and duodenal, pressures (Thomas and Kuntz 1926; Thomas and Crider 1935; Meschan and Quigley 1938; euigley and Werle 1940; Quigley et al 1942; Atkinson 1957; Andersson and Grossman 1965). 13

Concerns about the potential for physiological disturbance caused by the bulk of balloons, the length of the lumen sampled by large balloons and the excessive compliance of balloons which prevents accurate quantitative recordings of pressure changes, led to the development of miniature balloons filled with air (Atkinson et al 1957) or water (Anderson and Grossman 1965; Brink et al 1965). However, balloon sonsors do not detect adequately tonic and phasic elevations of pressure when these phenomena are superimposed (Dodds et al 1976), and are difficult to maintain in position within a narrow contracting zone due to their tendency to be squeezed proximally or distally. ln most studies, inilial localization of balloons at the gastroduodenal junction was achieved by assessing resistance to traction through the pylorus, and there was no feedback on assembly position during recording. Most of these disadvantages also apply to the cylindrical air-filled'pressure tonometer'of Thomas et al (1934) and the'optical manometer' used by Brody et al (1940). None of these techniques is appropriate for pyloric manometry.

2.3.1.2 lntraluminal Side-hole Manometry Recordings from side-holes perfused with water or saline via a low-compliance pneumohydraulic pump have been shown to reflect intraluminal pressures faithfully (Pope 1967; Dodds et al 1976; Arndorfer et al 1977; Valori et al 1986). Accordingly, this method of manometry has been widely accepted. The performance of systems incorporating side-holes in non-perfused water- or air-filled channels (Brody et al 1940; Atkinson et al 1957; Brink et al 1965) is unsatisfactory.

Side-holes are of course point sensors, and this limitation must be recognized when attempting to record from a narrow, mobile zone (Section 2.2'¡. For example, attempting to station a single side-hole at the gastroduodenal junction (Aste et al 1979; Pandolfo et al 1979; White et al 1981; Defilippi and Gomez 1985) is probably futile, irrespective of the method used' Wilhdrawal of side-holes across the gastroduodenal junction has been employed frequently to measure pyloric pressures, and this technique will allow brief recording from the narrow pylorus (Atkinson et al 1957; Andersson and Grossman 1965; lsenberg and Csendes 1972; Fisher and Cohen 1973; Fisher et al 1973; Kaye et al 1976; Valenzuela and Defilippi 1976; Valenzuela et al 1976; Aste et al 1979; pandolfo et al 1979; Phaosawasdi et al 1979; Mcshane et al 1980; Gaffney et al 1987). However, it may not be possible to properly identify the pylorus during these procedures, unless another technique such as potential difference measurements is used for localization (Andersson and Grossman 1g65; Fisher et al 1973) since, unlike the lower oesophageal sphincter, the pylorus is not usually tonically contracted in the resting state (Atkinson et al 1957; Andersson and Grossman 1965; Fisher and Cohen 1973) (see Sections 4.2.2 and 4.3'21' Furthermore, prolonged assessment of pyloric motility is not possible wilh these techniques, because of the practical difficulties of performing repeated pull-throughs in suitably small steps of catheter withdrawal. 14

The use of very closely spaced side-holes to measure pyloric motility is technically demanding. lf side-holes alone are used, they must be spaced less than 6mm apart on the manometric assembly, perhaps necessitaling six to eight side-holes spanning four to live centimetres, to allow both for the narrow zone of contraction and the axial mobility of the pylorus. Side-holes spaced at 9mm intervals would fail to rscord approximalely two thirds of localized phasic pyloric contractions (Heddle 1988b). Few studies have used such close sensors (Reynolds et al 1984 & 1985; Heddle et al 1988b), so many previous recordings of pyloric motilily with unanchored side-hole assemblies are likely to have under-recorded contractions localized lo the pylorus (Behar et al 1979; Biancani et al 1981; Mearin et al 1986; Bertiger et al 1987).

Although side-holes, and the sleeve sensor (see below), record changes in intraluminal pressure accuralely, low amplitude contractions which do not occlude the lumen may produce liltle or no rise in intraluminal pressure (Carlson et al 1966; Christensen 1987). Such contractions, even though fhey may be of mechanical importance, will therefore not be recorded by intraluminal manometry. Carlson et al (1966) demonstrated in fluoroscopic studies in dogs that some antral contractions, especially those ¡n the proximal antrum, do not result in lumen occlusion. Thus, while antral side-hole manometry records intraluminal pressures accurately, and provides much useful information, it does not detect all antral contractions. This limitation is also relevant to the proper identification of localized pyloric contractions (Chapter 4), and can only be overcome by the use of a more sensitive detector of non-lumen-occluding contractions (Chapter 8).

Finally, although side-hole manometric assemblies are of smaller diameter than early balloon sensors, the possibility that the technique itself alters gastrointestinal function must slill be considered (Heading 1980). Manometric catheter perfusate is unlikely to produce any physiological disturbance if it is not hyperosmolar, acidic or calorie-rich, and is infused at a rate slow enough to avoid luminal distension (Chapter 5). Potentially conflicting information is available on the effect of intestinal intubation on the rate of gastric emptying (Muller-Lissner et al 19g2; Longstreth et al 1975; Read et al 1983), and the effect of intubation on motility has nol been assessed. Better definition of the effecl(s) of intubation on gastric motor function is of fundamental importance to the proper interpretation of recordings of pyloric, antral and duodenal motility (Chapter 8).

2.3.1.3 The Sleeve Sensor First described by Dent over a decade ago, the sleeve sensor was originally designed for improved manometry of the lower oesophageal sphincter (Dent 1976). Since the sleeve sensor records pressure from any point along the length of its distensible membrane, it is ideally suited to recording from lhe very narrow pyloric zone while avoiding the logistic complexity involved with the use of multiple closely spaced side-holes (Maddern et al 1984; Allescher et al 1988b; 15

Heddle et al 1988a,bac; Houghton et al 1988aab; Treacy et al 1988aab; Fraser et al 1989aab). Combining a sleeve sensor for pyloric manometry wilh perfused side-holes or serosal strain gauges in the antrum and duodenum, enables prolonged and reliable recording of antro-pyloroduodenal motor patterns (Allescher et al 1988aab; Heddle et al 1988aab; Houghton et al l ggga&b). Sleeve/side-hole manometry, with assembly localization via continuous feedback from transmucosal potential difference measurements in the antrum and duodenum (Chapter 2.2.1), was used in all studies reported in this thesis.

There are several limitations of sleeve manometry. Firstly, the sleeve sensor, which protrudes into both the antrum and duodenum, records the maximal pressure from anywhere along its length of several centimetres (see above). Therefore, w¡th sleeve recording alone, it may not be certain that a non-localized pressure wave recorded by the sleeve involved the pylorus. Furthermore, it ¡s nof possible to determine whether a localized Pressure wave recorded by lhe sleeve was narrow enough to have involved only the pylorus, and not the adjacent terminal antrum and proximal duodenum as well. Understanding these considerations is critical for the correct identification of isolated pyloric pressure waves (Tougas el al 1987; Heddle et al 19BBa,bac; Houghton et al 1988aab) (Chapter 4). When interpreted in conjunction with side-hole recordings above and below the sleeve, localized and non-localized pressure waves can be distinguished with acceptable accuracy (Heddle et al 1988b). Using such an assembly, a localized pressurs wave recorded only by the sleeve can be assumed with reasonable justification to represent a lumen-occluding contraction generated by the pylorus. lt may, however, be useful to better characterize the topography of recorded waves, for example by the inclusion of a series of side-holes spaced along the sleeve. This deserves further evalualion (Chapter 8).

Other limitations of the sleeve sensor are related to its physical properties, such as its length and the compliance of its distensible membrane (Dent 1976). Membrane compliance results in pressure rise rates which vary along the length of the sleeve and which, except at its proximal end, are significantly less than those of perfused side-holes. Despite this, the sleeve records localized and non-localized prossure waves with good sensit¡vity (89% and 957" respectively) compared with side-holes (Heddle et al 1988b). ln addition, mechanical distortion of the sleeve membrane produces recording artefact, and renders pyloric tone measurements unreliable. This could occur, for example, by abutment of the sleeve sensor on the wall of the duodenal cap, or by angulation of the sleeve around the junction of the first and second parts of lhe duodenum. To limit this possibility, the sleeve must not be excessively long, and lhe catheter musl be flexible at the distal end of the sleeve. These design features were incorporated into the assemblies used for the studies reported in lhis thesis. The issue of whether gastrointeslinal inlubation alters motility remains controversial (Section 2.3.1.2, Chapter 8), but incorporation of a sleeve sensor inlo the manometric assembly is unlikely to have any additional effect (Heddle et al 1988b). 16

2.9.1.4 Endoscopy Direct visualization of the pylorus by endoscopy (Munk et al 1978; Aste et al 1979; Eyre-Brook et al 1983) has a number of disadvantages. Prolonged observation is not possible and endoscopy cannot give information about fed motility, since the presence of gastric contsnts obscures the mucosa and is dangerous due to the risk of aspiration. Endoscopic intubation, air insufflation and sedation could each interfere with normal motility patterns. Endoscopy has been used to position sensors such as impedance electrodes at the pylorus (Linhardt et al 1981; Eyre-Brook et al 1983; Johnson et al 1983), and initial localization across the gastroduodenal junction was probably adequate in these studies. However because lumen occlusion obscures vision, confirmalion of sensor localization at the pylorus would not have been possible during propagated contractions. Endoscopy, however, would be useful in the initial positioning of sensors if feedback from transmucosal potential difference measurements was subsequently used for sensor localization. Such an approach may be appropriate when studying patients with gastroparesis, ín whom there may be inadequate fasting antral contractions 1o permit transpyloric passage of a manometric assembly.

2.3.1.5 Measuremenls of Transpyloric Flow lnferences about pyloric motility have been made from measuring transpyloric flow in animals. Edin and co-workers (Edin et al 1979a; Edin 1980; Lidberg et al 1984) used such a technique in cals, however their acute surgical preparation involved antral and duodenal cannulae and ligatures which could have interfered with normal motility. A similar technique was used by Schulze-Delrieu and Wall (1983) in dogs, and Ruckebusch and Malberl (1986) in sheep. lsolation of the dislal pyloric muscle ring between rigid antral and duodenal cannulae is not possible without mechanical distortion of the pylorus. Furthermore, since these methods measure gastroduodenal resistance, rather than that of the pylorus alone, lhe effect on flow observed in these studies may result from a combination of terminal antral, proximal duodenal and/or pyloric segment contractions. Mearin et al (1986 & 1987) measured the resistance to a constant flow of air through a flaccid cylinder, positioned fluoroscopically across the pylorus in dogs. While this prsparation would not have caused much mechanical inlerference, there was no conlinuous feedback on sensor position, and there was the potential for artefactual elevations of airflow resistance due to bending of the cylinder over the gastroduodenal junction. With the exception of Mearin's approach, none of these techniques are applicable to humans.

2.3.1.6 lmpedance Eleclrodes Eyre-Brook et al (1983) studied pyloric closure using four silver wire elecrodes mounted on the surface of an intraluminal catheter. Luminal closure was registered as a fall in impedance measured between opposite pairs of electrodes when touched by the mucosa. The device was positioned endoscopically and maintained in position by a balloon in the duodenum. Doubts about 17

correct position during antral contractions remain, however, and the technique is unsuitable for prolonged studies or recording in the fed state.

2.3.2 Extralumlnal Techniques For Recording Pylorlc Motlllty 2.3.2.1Strain Gauges Strain gauges sutured to the serosa have been used in many important studies of pyloric motility in animals (Mir et al 1977 & 1979; Telford et al 1979; Prove and Ehrlein 1982; Allescher et al 19g8aab). Strain gauges record the occurrence of contractions which deform a small plate in the base of the device. The size and sensitivity of strain gauges differ from study to study, and there has been no satisfactory evaluation of the area of smooth muscle from which the strain gauge records. ln order 1o record specifically from the narrow pylorus, strain gauges must be very small, and must be located accurately at the level of the distal pyloric muscle loop.

Strain gauges, and induction coils (see below), have some advantages compared with manometry, since they can record low-amplitude non-lumen occluding contractions and, because they are positioned extraluminally, there is no potential for physical interference with transpyloric flow. Both techniques are suitable for prolonged recordings of fed or fasting motility in unanaesthetized animals, but they are clearly unsuitable for human studies. No studies have compared the recordings of strain gauges sited at the d¡stal pyloric muscle ring with simultaneous sleeve/side- :{ hole manometry. Such studies could provide direct evidence that localized increases in pyloric ll',1! j pressure recorded manometrically are indeed due to contraction of the distal pyloric muscle loop.

2.3.2.2 lnduction Coils Changes in the external diameters of the gastro¡ntestinal tract can be recorded in animals by induction coils (Brody and Quigley 1944; Louckes et al 1960; Ehrlein 1980; Ehrlein and Prove 1982; Prove and Ehrlein 1982; Keinke and Ehrlein 1983; Keinke et al 1984 & 1986; Ehrlein 1988). Although induction coils can be sutured to the serosa at the distal pyloric ring with some accuracy (Ehrlein 1988), they give only an indirect measure of pyloric contraction since pyloric diameters may also be influenced by other factors, such as traction from antral or duodenal

contractions, or pulsion f rom intraluminal contents. However, the motility pattêrns demonstrated in recent studies (Ehrlein 1988) suggest that lumen occlusion at the pylorus can be reliably recorded by induction coils if due attention is paid to sonsor localization.

2.3.2.3 Electromyography The identification and correct sampling of smooth muscle cells from ths distal pyloric muscle loop (Section 2.3.4) is the major unresolved methodological problem for electromyographic studies of the pylorus (for example, Bass et al 1961; Bortoff and Davis 1968; Edwards and Rowlands

! 18

1968; Fujii 1971; El-Sharkawy 1978; Papasova et al 1982). Current knowledge about pyloric electrical activity is discussed in Chapter 5.

2.3.3 Non-lnvasive Approaches For Recording Pylorlc Motility 2.3.3.1 Radiology Radiology has been used successfully to observe gastroduodenal motility in humans (Gershon- Cohen and Shay 1937; Torgersen 1942; Keet 1957; Williams 1962; Johnson 1975; White et al 1981; White et al 1983; Tougas 1987) and in animals (Cannon 1898; Carlson et al 1966; Prove and Ehrlein 1982; Ehrlein et al 1983; Seide and Ritman 1984; Keinke et al 1984 & 1986; Kumar et al 1987; Ehrlein 1988). Although these studies are hampered by an inability to accurately localize the pylorus on the images obtained, the lechniques are relatively non-invasive and can provide useful information about the pattern and timing of lumen- and non-lumen occluding contractions. Unfortunately, radiation exposure limits the use of fluoroscopy to animals or to brief periods of observation in humans. Furthermore, the introduction of non-physiological contrast media may conceivably interrupt the interdigestive migrating motor cycle, precluding the study of true fasting motility. Nevertheless, concurrent fluoroscopy and sleeve/side-hole manometry is likely to advance our understanding of the mechanical role of the pylorus. An example of this is the demonstration by Tougas et al (1987) in humans that localized pyloric contractions induced by intraduodenal infusion of lipid prevent lhe transpyloric flow of barium. .J -ilf

',1j 2.3.3.2 U ltrasonography Ultrasound is a non-invasive technique which does not involve ionizing radiation, and which avoids the introduction of non-physiological contrast media. lt may be an effective way of measuring gastric emptying and possibly antral contractions, but at present cannot convincingly record from the pylorus. The recording of pyloric contractions by ultrasound depends on imaging adjacent structures simultaneously in the same plane, and differentiating the pylorus from the antrum and duodenum. This presents a substantial challenge when studying the mobile, narrow gastrododenal junction, and the required expertise is not available in many centres. To date, all studies reporting successful pyloric imaging with ultrasound have come from the same group of researchers (Holt et al 1980; Adam et al 1982; King et al 1984, 1985 & 1987). Their conclusions regarding coordinated post-prandial motility of this region are consistent with fluoroscopic observations (Cannon 1898; Carlson et al 1966) and strain gauge/induction coil recordings (Ehrlein 1988). Ultrasound requires the presence of echogenic material, such as orange juice containing fine particles of bran (King et al 1984 & 1985), within the stomach for image production, and so is not suitable for use in the fasting state. Unlike manometry, however, non-lumen occluding contraclions. Future I ultrasound has the advantage of being able to record improvements in image resolution may also allow identification of discrete episodes of

I 19

transpyloric flow, enabling important correlations to be made between gastric emptying and patterns of antropyloroduodenal motility.

2.3.3.3 Electrogastrography Electrogastrography records electrical activity from the stomach via surface electrodes attached to the skin (Alvarez and Mahoney 1922; Smout et al 1980). Antral dysrhythmias have been described in patients with nausea and vomiting (You et al 1981), motion sickness (Stern et al 1987) and anorexia nervosa (Abell et al 1987), and disturbances in amplitude have been found after meals in patients with nausea and vomiting (Geldof et al 1986). Although electrogaslrography involves much complexity in signal processing and interpretation, it is simple and non-invasive for the subject. At present, however, electrogastrography cannot record specifically from the pylorus.

2.3.4 In Vitro Studies of Pyloric Muscle The mechanical and pharmacological properties of pyloric muscle have been investigated by studying isolated muscle strips from a number of species (Lipshutz and Cohen 1972; Fisher et al 1973; Anuras et al 1974; El-Sharkawy et al 1978; Golenhofen et al 1980; Ludtke et al 1983; Schulze-Delrieu and Shirazi 1983; Berk et al 1986). Ludtke et al (1983) opened resected specimens from the human gastroduodenal junction along the lesser curve to obtain strips of -t distal pyloric muscle loop. A similarly '|r,T muscle exclusively from the macroscopically-identified j precise approach was taken by Golenhofen et al (1980) in dogs. Although Schulze-Delrieu and Shirazi (1983) carefully dissected strips parallel 1o the lie of the fibres of the distal muscle loop, some of the pyloric torus, and thus possibly some fibres from the proximal pyloric muscle loop, was included_ in lhe muscle strips. More significant sampling problems may have occurred in other studies due to dissecting slrips oblique to the lie of the muscle fibres (Anuras 1974), or taking relatively wide muscle strips (Fisher et al 1973; Berk et al 1986).

2.4 CONCLUSTON

The narrowness and mobility of the pylorus each provides a substantial technical challenge for measurement of pyloric motor activity. Satisfactory recordings in humans can be achieved with careful localization of an intraluminal manomelric assembly, using continuous feedback on sensor position during recording. lntraluminal manomelric assemblies should incorporate a series of closely spaced side-holes or a sleeve sensor to monitor pyloric motility. t I l

I 20

CHAPTER 3

Measurement of Gastric EmPtying

3.1 INTRODUCTION Gastric emptying has been measured with numerous techniques, although few have been capable of assessing the pattern of emptying under physiological conditions. Concurrent measurement of gastric emptying and gastroduodenal motility provides invaluable informalion about the :t mechanical significance of various motor patterns. Few groups of investigators have successfully ll ''1! combined suitable measurements of emptying with accurate recordings of antral, pyloric or j duodenal motility (see, for example, Camilleri et al 1985; Houghton et al 1988aab; Heddle et al 1989). This section briefly describes the advantages and disadvantages of the most widely used methods of measuring gaslric emptying in humans (Table 3.1), particularly to explain the choice of a radionuclide technique to measure gastric emptying in conjunction with manometry in the studies described in this thesis.

9.2 INVASIVE TECHNIOUES FOR MEASURING GASTRIC EMPTYING IN HUMANS 3.2.1 lntubat¡on/Aspiration of Gastric Contents The capacity of intubation techniques lo measure fractional gastric emptying (Findlay el al 1972; Meyer 1987) underlies the substantial contribution they have made to understanding the physiology of gastric emptying in humans (Hunt and Spurrell 1951; Hunt 1956; Goldstein and Boyle 1965; George 1968). With the double sampling technique (George 1968), description of the time course of gastric emptying is possible in a single session, avoiding the repeated testing necessary in single aspiration studies, a constraint which resigned Hunt to missing breakfast on many occasions as he meticulously documented the behaviour of his own stomach in the 1950's. I

Ir Disadvanlages of intubation/aspiration lechniques include their complexity and invasiveness, which essenlially limit their use to research and make their concurrent use with manomelry impractical. The use of the concentration of a non-absorbable water-soluble marker 1o determine t 21

residual gastric volumes requ¡res the assumption of rapid equilibration of the marker within lhe gastric contents. ln addition, the instillation of liquid test meals directly into the stomach via a lube atters normal physiology. For example, receptive relaxation of the gastric fundus does not occur, and this may result in a faster initial rate of gastric emptying (Hunt 1956). Furthermore, as discussed previously (Section 2.3.1.2), intubation may itself affect gastric motility.

Emptying may also be measured by duodenal intubation, with constant perfusion of a non- absorbable dilution indicator and sampling downstream (Meeroff et al 1975). This method has the advanlage of being applicable for measurement of solid gastric emptying (Malagelada et al 1979). With this exceplion, however, aspiration techniques are best suited to measurements of liquid gastric emptying, as adequate aspiration of solid meals before triturition is usually not poss¡ble.

1. lnvasive Techniques . lntubation/aspiration of gastric contents

2. Non-invasive Techniques . Radionuclide labelling of meals . RadiologY . Ultrasound . Applied potential tomography . lmpedance epigastrograPhY . Absorption kinetics of orallY administered solutes

Table 3.1 Techniques for measurement of gastric emptying in humans.

3.3 NON.INVASIVE TECHNIOUES FOR MEASURING GASTRIC EMPTY¡NG 3.3.1 Radlology Radiological observations of gastric function have been employed since the studies of Cannon (1898). These techniques have the major disadvantage of substantial radiation exPosure. Although fluoroscopy is useful for determining patterns of contraction and recognizing gross degrees of gastric relention (Cannon 1898; Carlson et al 1966; Tougas 1987), it cannot be used to measure gastric emptying accuralely, since the amount of barium remaining within the stomach, or the volume of gastric effluent, cannot be determined. Furthermore, liquid barium sulphate is physically very dense and non-nutrient. Contrast radiography also cannot properly measure the emptying of solid food, a deficiency which some investigators have addressed by

þ 22

combin¡ng barium with an ingested meal (Cannon 1898; Carlson et al 1966; Wilkinson and Johnson 1973). Feldman et al (1984) recently reported the use of radio-opaque markers to study gastric emptying during fasting. A single plain X-ray of the abdomen taken 6 hours after ingeslion of lhe markers was reported to be a sensitive method of diagnosing delayed gastric emptying. While this technique may be useful, it is likely to only assess the integrily of the antral component of the interdigestive migrating molor complex, which is absent in some types of gastroparesis (Malagelada et al 1980; Camilleri and Malagelada 1984).

3.3.2 Ultrasonography Real-time ultrasound scanning is safe, comfortable for the subject and is an attractive alternative to radiology or scintigraphy because it does not use ionizing radiation. Ultra- sonography can be used to measure gastric volume indirectly. Repealed measurements can provide gastric emptying curves. Since the volume of gastric secrelions remains unknown, fractional gastric emplying cannot be measured with confidence. Although the technique has been used succsssfully to study gastric emptying of liquid meals in humans (Bateman 1982), it has not gained wide acceptance. Ultrasound cannot measure the emptying of most solid meals accurately. ln addition, poor visualization in obese subjects and difficulties wilh imaging because of a high transverse-lying stomach can pose significant problems (Mamtora and Thompson 1989).

3.3.3 Applied Potential Tomography and lmpedance Epigastrography These recently described methods use changes in electrical impedance or resistivity to measure gastric emptying (Avill et al 1987; Gilbey and Watkins 1987). Both are simple, relatively cheap, non-invasive and do not require the use of radiation. Profiles for liquid gastric emptying obtained by applied potential tomography compare favourably with scintigraphy or intubation/aspiration studies, however impedance epigastrography is somewhat less accurate (Mangnall 1989). Neither technique is able to measure solid gastric emptying reliably and both (especially applied potential tomography) require premedication with an H, antagonist such as cimetidine (ranitidine increases the rate of gastric emptying) to prevent fluctuations in gastric impedance due to secretion of acid during emptying studies. Nevertheless, there may well be a place for these non- invasive tests in the future, although concerns about measurement accuracy remain.

3.3.4 Absorptlon KinEtics of Orally Administered Drugs The absorption of orally administered drugs, such as paracotamol, depends on the rate of delivery of the drug to the small intestine. Blood or salivary concentrations of absorbed drug products have been used as an indirect measure of the rate of gastric emptying (Nimmo 1976), but this technique is probably too inaccurate for widespread application. Plasma drug levels may be dependent on other variable factors, such as the speed and completêness of mucosal absorption, as well as the rate of gastr¡c emptying. The method can only be used to measure liquid gastric emptying. 23

3.3.5 Radlonucllde Technlques Radionuclide labelling of ingested meals offers a non-invasive and accurate means of quantitating gastric emptying (Figures 3.1 and 3.2). Scintigraphic measurements are sensitive and reproducible, and are thus widely used in research and clinical medicine. They are considered to be the best available technique for measuring gastric emptying (Read 1989). Because they are non-invasive and comfortable for the subject, radionuclide measurements of gastric smptying are particularly suitable for use in humans combined with simultaneous intraluminal manometry. Gastric emptying studies reported in this thesis all employed measurements of the gastric emptying of a standardized isotopically-labelled solid meal (Chapter 7).

3.3.5.1 Radionuclide Markers First described over 20 years ago (Griffith et al 1966), initial radionuclide techniques were either unable to distinguish between radioactivity in the stomach and the small intestine (Bromoten et al 1968), or took an unacceptably long t¡me to acquire data (Heading et al 1971). These deficiencies were partly due to the limitations of early scintiscanners, as well as the properties of the isotopic labels available at the time. Particular properties of radio-isotopic markers are needed for reliable and reproducible measurements of gastric emptying. The ideal radiopharmaceulical should be non-toxic, non-absorbable, homogeneously distributed throughout lhe meal, cheap and give a low radiation burden. Since the liquid and solid phases of a meal empty at different rates, it is essential that virtually all the marker remains in the phase it is designed to measure during gastric emptying (Horowitz et al 1985). Adherence of a liquid marker to solid food leads to underestimation of the rale of liquid emptying (Heading et al 1976). Conversely, u'Cr weakly adherent solid food markers, such as as chromate (Griffith et al 1966), dissociate into liquid gastric contents, and result in overestimation of the rate of solid emptying. slCr Subsequent studies indicated that as much as 90% of the used by Griffith to label scambled egg could have dissociated from the semi-solid component of the meal (Meyer et al 1976).

ttt'ln tl1'ln-DTPA, There are now a number of radiopharmaceuticals, such as and which are suitable for measuring the gastric emptying of liquid meals (Heading et al 1971). The development of adequate markers of the rate of gastric emptying of digestible solid meals has been more ee'Tc-sulphur difficult. The best solid marker is colloid, which can be incorporated into chicken liver by intravenous injection into the chicken 30 minutes before the chicken is killed (Meyer et se'Tc-sulphur al 1976). colloid may also be combined in vitro with fresh chicken liver or eggs, ee'Tc-tin with only slightly less labelling efficiency than the in vivo technique. incorporated inlo tttl-labelled pancakes may also be satisfactory, as may incorporation of resin pellets 0.5- 1.8 mm in diameter into a mixed-nutrient solid meal (Camilleri et al 1989).

131land ee'Tc Labelling of fibre and bran, with respectively, allows assessment of the transil of 75Se-glycerol non-digestible solids. Recently, it has been reported that triether butter is a 24

Flgure 3.1 eem after ingestion, from a A) Scintigraphic image of a Tc-labelled solid meal in the stomach, 5 min margin has posteriorly positioned gamma camera. Counts are concentrated in the fundus of the stomach' A screen with a cursor, been constructed around the lotal stomach "region of interest' on the computer subject's back enabling selective quantification of gamma images from this area' The marker taped to the (indicated by the cross) allows correction for subject movement' B) The gastric image can be divided into próximal and distal regions of interest. These sequenlial stomach, and scintigraphic images show the passage of a solid meal from the proximal to the distal subsequently into the small intestine. 25

su¡table label for fat (Jian et al 1982), but this isotope delivers a high radiation dose and is not readily available. Although it has been stated that solid phase studies of gastric emptying may be more accurate in patients with gastroparesis than measuremênts of liquid emptying (Malmud et al 1992), this may not be the case if a liquid phase is nutrient-rich and therefore empties more slowly. Nevertheless, if only one phase is to be studied, a solid marker may be preferable (Camilleri et al 1985; Horowitz and Akkermans 1989).

3.3.5.2 Other Methodological lssues The rate of gastric emptying is substantially influenced by the composition and volume of thê test meal. Half-emptying times for solid meals are known to increase with meal size (Meyer et al 1976; Çhristian et al 1980). Meyer et al (1976) reported markedly different half-emptying times of 170 minutes and 80 minutes for 8509 and 4259 solid meals respectively, but the emptying of these otherwise identical meals in fact proceeded at a similar rate of about 2 g/min. Thus, the amount emptied/min may depend more on the caloric composilion of the meal (Brener et al 1983; Chapter 5), rather than its volume alone'

Viscosity, particle size and particle density are also important determinants of the rate of emplying of solid and semi-solid meals (Meyer 1987). lncreased particle size necessitates more prolonged grinding and results primarily in lengthening of lhe lag phase (Figure 3.2) before gastric emptying begins. Gastric emptying of semi-solids may be slowed if the viscosity of the gastric contents is high enough (Prove and Ehrlein 1982). The retardant effect of increased particle density on the gastric emptying of nutrients may occur by virtue of a greater calorie load per unit volume, but the influence of density may be more subtle. For example, the density of food particles affects the behaviour of those particles in liquid suspension (see also Section 6.3.2). Thus, when ingested simullaneously with a solid meal, plastic spheres with a density of 2.0 (which tend to settle out of suspension) or 0.5 (which float), empty more slowly than spheres with a density of 1.0 (Meyer et al 1985). For clinical studies of gaslric emptying, a meal size and composition should be chosen so that emptying proceeds over a convenient time span of perhaps several hours. The meal should be palatable and able to be ingested with comfort over a relatively short period of time. Standardization of the test meal is essential.

Gravity may also play a small but significant role in determining the rate of gastric emptying. There is some evidence that non-nutrient liquids empty more rapidly when the subject is erect (Hunt et al 1965). The active control of nutrient liquid emptying by motor mechanisms (Seclion 6.3.1) results in posture having little effect, except when these mechanisms are altered, for example by vagotomy and pyloroplasty when nutrients emPty more rapidly in the erect posture (Gulsrud et al 1980). Recently, it has been suggesled that the gastric emptying of solids may also be faster in upright subjects (Moore et al 1988). 26

Lag Phase v7777777m! 100 a-oto .'-.t' a o t-.aa. O Solld 80 o O Liquld ¡o oo o a a a o ¡! a oo da RETENTION % oo - o daa 40 æ oo oo

a o

o 0 20 40 60 80 100 120 POST CIBAL MINUTES

Figure 3.2 Gãsfic emptying curves for a dlgestible solid (eem Tc-chicken liver/ground beef) and liquid (10% dextrose phase labelled *¡{.r ira. ln-DTPA) in a normal subject. The pattern of solid emptying is characterized by a lag before food leaves the stomach, followed by a linear emptying phase. Emptying of liquid is more rapid lhan that of the solid, and has a non-linear pattern wilh a slope that decreases with time. 27

Finally, the time of the day at which the meal is ingested may be important' Goo et al (1987) reported a circadian variation in the rate of solid gastric emptying, with the rate being slower in lhe evening (see Section 9.4 for further detail). These issues require clarification, but emphasize the need for standardization of subject position and the time of meal ingestion.

3.3.5.3 Data Acquisition and Processing lmproved resolution of detected isotopic counts resulted from the introduction of the gamma camera (Harvey et al 1970). Linking the gamma camera to a computer permits rapid determination of the distribution of radioactivity wilhin the abdomen, a facility essential for the accurate estimation of the lag phase for solid gastric emptying, and of the initial rate of emptying, which may be rapid especially for isotonic liquids or after partial gastrectomy (Heading ef al 1976; Brener et al 1983). From computer-generated images the stomach can be identified on a screen by its anatomical shape, around which a 'region of interest" can be drawn wifh a cursor, excluding the small intestine (Figure 3.1a). Recently, Collins, Horowitz and co- workers developed a method for determining proximal and distal gastric regions-of-interest (Figure 3.1b), enabling estimation of the regional distribution of a meal within the stomach, as well as the rate of gastric empty¡ng (Collins et al 1983). By analysing changes in radioactivity within small regions of interest drawn over the antrum, an indirect assessment of antral contractions can be obtained (Akkermans et al 1980; Stacher et al 1984). With more se'Tc 1"ln¡ sophisticated gamma cameras, radionuclides of different energies 1eg and may be studied simultaneously, enabling concurrent measurement of solid and liquid smptying (Heading et al 1g76). Normal gastric emptying curvos obtained with a radionuclide technique are shown in Figure 3.2. Parameters used to dsscribe the characteristics of emptying curves have been previously discussed by Dugas (1982) and Elashoff (1983) (Figures 3.2 and 7.3; Section 7'5'2).

3.3.5.4 Geomelrical Errors of Radionuclide Methods The accuracy and reproducibility of modern radioisotopic measurements is dependent on the recognition and correction of a number of methodological variables, including geometrical errors in data acquisition (Christian et al 1980 & 1983). Appropriate allowance must be made for patient movement, radionuclide decay, and attenuation of gamma rays resulting from the changes in the distribution of the meal within the stomach during emptying studies (Collins et al 1984). After ingestion, food initially moves anteriorly within lhe stomach, so a gamma camera positioned posteriorly will overestimate the rate of emptying as the food moves away from the gamma camera, whereas an anterior detector will underestimate the rate of emptying (Heading 1976; Collins et al 1983). Resultant errors in the measurement of the time for 50% emptying nn'Tc-labelled tl3-ln-labelled have been estimated to average 2r,/" lor solids, and 18% for liquids. ln many previous studies, no correction was made for attenuation (Meyer et al 1976; MacGregor et al '1977; Lavigne et al 1978). This problem has been addressed by others, 28

however, by combined anterior/posterior detection to obtain geometric mean counts, either by rotating the subject or using two cameras (Tothill et al 1978; Moore et al 1980), or by the application of correction factors derived from a lateral image of the stomach (Collins et al

1 e83).

Deflection of radiation by body tissues, or nscaltêr", may result in degradation of lhe scintigraphic image. For example, when two isotopes are used a proportion of scattered radiation from the higher energy isotope reaches the detector as lower energy pholons, indistinguishable from those emitted from the lower energy isotope. lnterference ¡nto both low and high energy *windows' may arise if the two isotopes are similar. The contribution of this interference, or 'Compton scatter', can be minimized by using small amounts of high energy isotope and by the use of correction factors (van Deventer et al 1983). ln addition, some gamma rays from the stomach may snter the detector at an angle, passing through, instead of between, the parallel lead septae of the detector. These rays may be detected outside the gastric region-of-interest, resulting in reduced image quality and underestimation of intragastric activity (Horowitz and Akkermans 1989). The amount of this 'septal penetration" remains fairly constant throughout the study, so if the counls in the gastric region-of-interest at the lime of meal completion are taken as lOOVq its effect on analysis will probably not be significant (Loo et al 1984).

3.3.5.5 Limitations of Radionuclide Methods One limitation of radionuclide techniques is that the volume of gastric secrelion is unknown, so that isotopic markers become progressively diluted to an unknown degree. The effect of gastric secretion is probably greatest on liquid markers. Radioisotopic techniques are suitable for prolonged and accurate monitoring of gastric emptying of normal solid, liquid and mixed solid- liquid meals, but concerns about about radiation exposure preclude their use in pregnancy.

3.4 CONCLUSTON Numerous techniques have been used for the measurement of gastric emptying in humans. Radio- isotopic gastric imaging techniques at present provide the most accurate and clinically applicable method to quantify gastric emptying in health and disease. These methods have the additional important advantage of being non-invasive and comfortable for the subject, and are especially suitable for concurrent use with intraluminal recordings of gastric motility. The radiopharmaceuticals used must have a high labelling efficiency and stability, and correction techniques should be used to compensate for errors such as radionuclide decay, radionuclide gamma ray attenuation and, if two isotopes are used, Complon scatter. 29

CHAPTER 4

Coordinated Patterns of Pyloric Motiltty

4.1 INTRODUCTION There is limited knowledge about the normal patterns of pyloric motility. However, intuition suggests that the activity of the antrum, pylorus and duodenum must be interrelated to enable their efficient function, and that the close regulation of gastric emptying should therefore involve coordinated motor activity. Doubts about whether some palterns of contraction are coordinated or not have arisen partly because of uncertainties about what "coordination" really is, and an appropriate and generally accepted definition of "coordination" is still required. ln addition, however, many early studies which examined the relationships between pyloric, antral and duodenal motility employed inappropriate methodology (Chapter 2), and the recent intraluminal manometric studies which described the pattern of localized pyloric motility more accurately must still be interpreted in the light of the known limitations of this technique lo record antral contractions. Pyloric motor patterns have been evalualed with technically adequate methods under only a limited variety of intraluminal conditions, and further studies are essential to understand more clearly the mechanical role of the pylorus (Chapter 6).

This chapter will discuss an appropriate definition for the term "coordination", then critically examine the current state of knowledge about the patterns of pyloric motility, with particular attention to human studies. Although patterns of motility in fasting and fed states have some similarities, mechanical effects under these conditions may be quite different. For example, although the numbers of antral contractions during phase lll of the interdigestive migrating motor cycle are similar to those observed in the fed state, phase lll aclivity is associated with the expulsion from the stomach of large particles, which are retained during normal gaslric emptying. For this reason, fasting and fed patterns of motilily will be considered separately. 30

4-1.1 Dellnltion of Antropyloroduodenal Coordinatlon The term "antroduodenal coordination' has been used frequently in the past to imply 'the characteristic pattern [of antral and duodenal motility] emerging after the intake of food' (Thomas and Crider 1935; Allen et al 1964; Schuurkes and van Neulen 1984). This concept of coordination should be regarded as incomplete. ln this thesis, 'coordination" will be used to refer to 'lhe lemporal association of antral, pyloric and duodenal contractions, or pressure waves, within one cycle of electrical control activity.. Not only does this definition include the pylorus, but it allows the presence of coordination in both fasting and fed states. Furthermore, it incorporates the concept of inhibition of motility. For example, it will be argued that the temporal association of localized pyloric contraction with suppression of antral and duodenal contractions (Sections 4.2.2 and 4.3.2) should be regarded as a pattern of coordinated motility. The motor function of the gastric fundus is also likely to be coordinated with that of the pylorus, antrum and duodenum (Azpiroz and Malagelada 1985aab; Meyer 1987), but simultaneous measurements of fundal tone with antropyloroduodenal motility are required to clarify this issue. Under the terms of the above definition, the different phases of the interdigestive migrating motor cycle are not considered to be an example of coordination, nor is any influence of the central nsrvous system or gastrointestinal hormones on pyloric motility (Chapter 5). These factors could be regarded as 'regulaling", rather than "coordinating", pyloric motor function.

Finally, whether pyloric pressure waves are judged to be associated, and therefore potentially coordinated, with those of the antrum and duodenum, depends on the application of rules of relative timing which remain partly intuitive (for example, see Seclion 4.2.1). The need for this arises because the relationships being assessed are as yet incompletely understood. The rules used in the studies reported in this thesis (Chapter 7) are based on manometric observations of normal motility by myself and others (Heddle et al 1988a,bac; Houghton et al 1988aab), as well as the known propagation velocity of smooth muscle electrical control aclivity (Chapter 5). Differences in the timing of coordinated contractions reported belween studies may be due in part to the variability of normal motor patterns (Table 4.1), but methodological factors may also be important. For example, radiological pyloric closure occurs prior to peak pyloric contraction or wall movement recorded by strain gauges and induction coils respectively (Ehrlein and Hiesenger 1982; Seide and Ritman 1982). ln addition, because the onset of a pressure wave recorded manomelrically occurs only when the contraction occludes the lumen (Chapter 2), the timing of prsssure waves recorded at different levels of the gut may differ from the timing of electrical activity or muscle contractions, and may vary depending on the resling diameter of the lumen.

4.2 COORDINATED PYLORIC MOTILITY DURING FASTINq There has been considerable disagreement in the past about the presence or absence of coordination during fasting. A number of technically unsatisfactory studies have, perhaps 31

fortuitously, described a temporal association between pyloric, antral and duodenal activity (Wheelon and Thomas 1922; Meschan and Quigley 1938; Pandolfo et al 1979). However, the results of studies using techniques which were probably sufficient to record from the pylorus, including radiology, strain gauges and induction coils, and side-hole manometry (Chapter 2), also indicate that patterns of fasting motility may frequently be coordinated (Brody et al 1940; Smith et al 1957; Louckes et al 1960; Carlson et al 1966; While et al 1981; Ehrlein and Hiesenger 1982; Keinke and Ehrlein 1983; Keinke et al 1984; Ehrlein 1988). More importanlly, lhe presence of coordination during fasting is supported by recent detailed sleeve/side-hole recordings of pyloric motility in humans (Heddle et al 1988a,bac; Houghton et al 1988aab).

The absence of coordination reported by some investigators (Allen et al 1964; Ehrlein and Heisenger 1982; Remington et al 1983; Schuurkes and van Nueten 1984; Defilippi and Gomez 1985) is probably due to problems of definition (above) as well as sensor localization (Chapler 21. For example, the use of a single side-hole stationed at the gastroduodenal junction by Defilippi and Gomez (1985) suggests that satisfactory pyloric recordings were not achieved in that study.

4.2.1 Patterns of Propagated Motility During Fasting Most studies indicate that, during phases ll and lll of the interdigestive migrating motor cycle, the pylorus normally contracts allef the antrum and þg[Otg the duodenum. Pyloric Pressure waves which occur after lhe terminal antrum were convincingly demonstrated in the fasting sleeve/side-hole manometric studies of Heddle et al (1988b) and Houghton et al (1988aab). Coordinated palterns of motility were considered by these authors to be dominant during fasting activity. Examples of the most common of these patterns are shown in Figure 4.1 and Table 4.1. These authors considered pressure waves to be coordinated if a single contraction, or clusters of 2 to g contractions, occurred within 10 seconds in at least three separate, but not necessarily adjacent channels, and were separated by quiescence lasting 20 seconds, equivalent to one cycle of the antral contractile frequency (note the comments in Section 4.1.1). Propagation of a contraction was said to have taken place when the leading edge of the contraction complex occurred >1 sec, but <5 sec after a similar contraction in the adjacent aborad recording site.

An elegant description of fasting motor patterns was provided by Wheelon and Thomas (1922). These aulhors observed the pylorus to be relaxed as the antral wave approached, and then to contract in sequence, beginning when the antral contraction was al its height, reaching its maximum when the antrum was relaxing and the initial duodenal contaction was beginning, and then relaxing as the contraction passed down the duodenum. While these studies may accurately reflect the relative timing of antral and duodenal contractions, and in this regard are consistent with the findings of Houghton et al (1988aab) cited above, the balloon sênsors used in their studies would not have reliably recorded from the pylorus (Chapter 2). 32

7 ANTRAL pH 4 I 7 DUODENAL PH 4 ! 80 8oo 0 80 GASTRIC PRESSURE 7rc (mmHg) 0

80 6m o 80 PYLORIC PRESSURE 5¿o (sleeve, mmHg) 0 80 4rc

80 3oo DUODENAL PRESSURE 0 (mmHg) 80 ¿40 0 s0 1¿o 0

I Minute

Figure 4.1 Antral, pyloric and duodenal pressure activ¡ty under fasting conditions, showing antropyloroduodenal coordination and its association with antral and duodenal pH changes. Pressures are measured from a 4.5 cm long sleeve sensor located across the pylorus (channel 5), and from side-holes situated 0, 3 and 6 cm above (channels 6, 7 and I respectively) and 0, 3, 6 and 9 cm below the sleeve (channels 4,3,2 and 1). Antral contractions are associated with 1 to 4 duodenal contractions. Pyloric motil¡ty recorded by the sleeve sensor during propagaled contractions cannot be separately distinguished from terminal antral and proximal duodenal motility. (Adapted from Houghton et al 1988b, with permission). 33

Fasting period Solid lag period Solid emptying period (h-1) (h- t¡ (h-1 )

Coordinated evenls Total 1o (3-60) 51 (7-14s) 63 (34-148)a Antroduodenal 1o (0-60) 34 (o-82) 42 (18-s7) Antral 0 (0-26) 5 (o-66)a 14 (3-t O1¡a'u Pyloroduodenal o (o-12) 5 (2-18) 5 (o-12) Duodenal o (0-4) 1 (0-6) o (o-4) lnvolving the antrum 0 (0-60) 3s (0-142) s8 (24-147) lnvolving the proximal antrum 1o (0-60) 3 (o-42) 30 (10-102)b

Showing antral propagation 6 (o-38) 14 (0-so) 27 (8-e7) Showing duodenal propagation 12 (o-30) 31 (6-54) 35 (1s-44)a lsolated pyloric pressure waves 0 (0-44) 7 (o-s2) 22 (10-64)a

Table 4.1 Pressure wave patterns recorded by sleeve/side-hole assembly during fasting and the emptying of the solid component of a mixed meal containing chicken liver/ground beef and normal saline. Results are median and b range (n=B). a Statistically significant difference from the fasting period (p<0.05). Statistically significant cjiíference from ihe soiici iag period (p<0.05). (From Houghtorr eî al i988a, wiih permission.)

It is likely that, under different conditions, the timing of pyloric lumen occlusion varies relative to antral and duodenal contraction, resulting in a wide variety of patterns of coordinated pyloric motility (Table 4.1). Detailed side-hole recordings also indicate that antegrade, retrograde or synchronous pressure wave patterns occur, over varying luminal lengths (Heddle et al 1988b; Houghton 1988a). Synchronous contractions of the pylorus and the terminal antrum have also been described in studies using fluoroscopy, ultrasonography, strain gauges and induction coils (Torgersen 1942; Keet 1957; Carlson et al 1966; Ehrlein and Hiesenger 1982; King et al 1984).

The temporal association of pyloric contraction with one or a variable number of 1 to 6 duodenal

contractions has been demonstrated in a number of studies (Carlson et al 1966; White et al 1981; Ehrlein and Hiesenger 1982; King et al 1985). Coordinated pyloric and duodenal contractions may occur in the absence of detectable conlraction in the antrum (White et al 1981;Eyre-Brook et al 1983; Houghton et al 1988a&b). lnstances of pyloric closure occurring after duodenal contraction in dogs have also been reported (Eyre-Brook et al 1983; Ehrlein and Heisenger 1982).

ln anaesthetized dogs, electrical field stimulation of intramural antral nerves is associated with abolition of localized pyloric contractions (Allescher et al 1988b). Whether this inhibitory neural mechanism is important in the normal control of fasting human motility is unclear. A phase of 34

pyloric inhibition or relaxation ahead of the approaching antral wave was described in several early studies (Wheelon and Thomas 1922; Meschan and Quigley 1938), but lhis was probably either artefactual, due 1o displacement of the balloon away from the gastroduodenal junction, or simply a reflection of 'the negative phase of the rhythmic contraction cycle'(Thomas 1957).

4.2.2 Locallzed Pylorlc Motlllty Durlng Fastlng Recent data suggest that localized pyloric contractions sometimes occur during fasting (for example, Allescher et al 1988b; Heddle et al 1988a; Houghton et al 1988a). ln fasted humans, lsolated pylorlc pressure waves, or lPPWs, are recorded manometrically over a very short luminal segment, similar to the IPPWs stimulated by intraduodenal infusion of lipid (Section 4.3.2 and Figure 4.4). The invariable association of IPPWs with suppression of distal antral and proximal duodenal contractions, suggests that this pattern also represents an example of coordination, in this case of pyloric excitation with antral and duodenal motor inhibition. Because the manometric definition of IPPWs depends on concurrent side-hole measurement of antral and duodenual motility, the completeness of suppression of antral contractions implied in these descriptions remains in doubt (Chapter 2). ln anaesthetized dogs, concurrent manometric/strain gauge studies indicate that, at normal recording sensitivity, sleeve-recorded IPPWs occur almost exclusively without any indication of antral or duodenal contraction (Allescher et al 1988b). At high gain settings, however, small antral and duodenal contractions were detected, suggesting that the definition of isolated pyloric contractions may be relative, and dependent on the amplification used in the measurement of antral and duodenal motility (Chapter 8). ln humans, isolated pyloric pressure waves (lPPWs) are infrequent ¡n the fasting stale (see, for example, Table 4.1). IPPWs have been observed interspersed amongst the propagated patterns of antropyloroduodenal molility described above (Houghton et al 1988a). Spontaneous IPPWs may occasionally interrupt phase 1 of the interdigestive migrating motor cycle (Heddle et al 1988b). Localized phasic contractions have been described in other studies using stationed or withdrawn side-holes (Fisher and Cohen 1973, Kaye et al 1976; Pandolfo et al 1979; Aste et al 1979; White et al 1983), or intraluminal balloons (Quigley et al 1942). These latlar techniques did not allow prolonged pyloric manometry (Chapter 2), so while it is possible that the local events they recorded were indeed generated by the pylorus, their incidence would have been underestimated.

Tonic contraction of the pylorus also occurs only rarely during fasting. ln a detailed study, Heddle et al (1988b) recorded only transient elevation of fasting basal pyloric pressure, consistent wilh other technically adequate manometric observations (Tougas et al 1987; Heddle et al 1988c; Fraser et al 1989aab). The elevation of fasting basal pyloric pressure reported by Houghton et al (1988b) may have been artefactual, due to distortion of the sleeve sensor by antral and duodenal pH probes attached to the manometric assembly in that study. ln dogs, 35

inductographic measurements suggest that lhe pylorus is usually open during phase I of the migrating motor cycle. ln contrast, tonic pyloric contraction has been observed during fasting in chloralose-urethane anaesthetized dogs, where fasting sleeve-recorded pyloric motility is frequently characterized by spontaneous IPPWs and high resting tone (Allescher et al 1988b). These results must be interpreted with caution, as they may have been influenced by general anaesthesia (Chapter 5) and local manipulation.

4.3 COORDINATED PYLORIC MOTILITY IN THE FED STATE ln terms of digestive significance, the patterns of coordination which occur after ingestion of food are more important than those observed during fasting. Although there have been few satisfactory studies of fed pyloric motility, the weight of evidence suggests that the activity of the pylorus is coordinated with that of the antrum and duodenum after the ingestion of food in many species, including humans, monkeys, dogs, cats and pigs (Allen et al 1964; Bortoff and Davis 1968; Adam et al 1982; Heddle et al 1988a; Houghlon et al 1988aab; Treacy et al 1988).

4.3.1 Patterns of Propagated Motlllty ln the Fed State Most technically adequate studies suggest that lhe predominant pattern of fed pyloric motility in humans is the sequential contraction of the pylorus after the antrum and þE[g¡g the duodenum, so that these regions apparenlly function as a single unit involved in peristaltic activity. Sleeve/ side-hole manometric studies have demonstrated frequent coordinated, aborally propagated events (Table 4.1, Figure 4.2) during emptying of solids and nutrient or non-nutrient liquids (Heddle et al 1988d; Houghton et al 1988aab). lt must be recognized, however, that the sleeve sensor, used to monitor pyloric motility in these studies, does not manometrically distinguish pyloric prsssure waves from terminal antral or proximal duodenal pressure waves during propagated pressure wave activity (Section 2.3.1). l1 remains an assumption, therefore, albeit a reasonable one, that the pylorus participated in these contractions, and there is a need for a detailed evaluation of the topography of pressure waves sweeping across the pyloric segment in the fed state, equivalent to that done during fasting (Section 4.2.3) or after intraduodenal lipid infusion (Section 4.3.3). Propagated contractions were also recognized in humans in early fluoroscopic studies (Mclure et al 1920; Cole 1928; Thomas 1957) and more recently by real- time ultrasonography (King et al 1984, 1985 & 1987), although there are some concerns about the adequacy of pyloric imaging using these techniques (Chapter 2). King and co-workers studied the gastroduodenal region after the ingeslion of 500 ml of dilute orange juice, containing 0.59 of echogenic bran particles. Closure of the pylorus was observed to occur at the midpoint of terminal antraf wall movements (lumen and non-lumen occluding), as peristaltic waves, occurring at a rate of 2.0 to 3.3/min, reached the pylorus. The duration of antral conlractions was approximately 4 sec, lhe latter half of which coincided with pyloric contraction, reproducing the 'lerminal antral contraclions" of Code (Carlson et al 1966) described below. The majority of 36

GâSIRIC PRftSUË lmmHgl 7

PYLo¡lc trtËËr PnfSSURt 5 (rlntlgl l I

Dl.þDttrA[ PÉSS UnE ldí*lgl 2- 1o

I illnul.

Figure 4.2 Antral, pyloric and duodenal motility in a healthy volunte€r during emptying of the solid component (chicken liver/ground beef burger) of a solid-liquid mixed meal. Pressures are recorded by a 4.5 cm long sleeve sensor across the pylorus (channel 5), and antral and duodenal side-holes situated 0, 3 and 6 cm above (channels 6, 7 and 8) and O, 3 6 and 9 cm below the sleeve (channels 4,3,2 and 1) respectively. There are frequent antropyloroduodenal and antropyloric coordinated pressure waves. (From Houghton et al 1988a, with permission.)

GASIRIC PRESSUNt 7o lmntlgl o D 6

PltoRlC (rlæw) 5 PRESSUf,t {nntþl a

3 DUODTMT PËSSUE {nnllg) 2

I Mlnutc

FIgure 4.3 Sleeve/slde-hole manometr¡c tracing of pyloric motility in a healthy volunteer during the emptying of the liquid dextrose component of a solid-liquid mixed meal conteining 25/" dextrose logether wilh a 1009 chicken liver/ground beef burger. Antral, pyloric and duodenal pressures are recorded from the same sites as those described in Figure 4.2. Motility is dominated by isolated pyloric pressure waves (lPPWs). (From Houghton et al 1988a, with permission.) 37

pyloric conlraclions (67%) was also associated with one or two proximal duodenal contractions, 94I" of which occurred about 1 sec after pyloric closure (range 1 sec before lo 2 sec after).

The results of animal studies are broadly consistent with those in humans. Ehrlein (1988), using induction coils and strain gauges located specifically at the level of the distal pyloric muscle ring, reported that pyloric contractions all began within 4 sec of maximal antral contraction, and reached maximum contraction 2.2 søc after the peak contraction of the antrum 2.5 cm proximally. No synchronous contractions of the antrum and pylorus were observed. These observations are supported by several other similar studies (Brody and Quigley 1944; Louckes et al 1960; Ehrlein and Hiesenger 1982; Prove and Ehrlein 1982), in which the location of the sensor, however, was less clearly indicated. Fluoroscopy during inductographic studies also suggests that the pylorus opens and closes in the rhythm of antral waves, being at ils widest diameter as contractions travel over the mid-antrum (Louckes et al 1960; Ehrlein 1988). Meticulous observations by Carlson et al (1966), during the emptying of a barium-containing meal in dogs, showed that as the peristallic wave passed down the antrum, the pylorus appeared to close simullaneously with the terminal antrum, followed in 85% of cases by duodenal contractions. Daniel (1965) noted radiologically that the sxtent of a single antral wave in dogs increased from 4 cm in the antrum to 15.5 cm in the terminal antrum, an observation which may be explained on electrophysiological grounds (Chapter 5).

4.3.2 Locallzed Pyloric Motillty in the Fed State Limited fluoroscopic observations during the emptying of nutrient meals or during intraduodenal nutrient infusions (Louckes et al 1960; White et al 1983; Fisher and Cohen 1973; Tougas 1987), have demonstrated the occurrence of phasic contractions ovor a very short luminal length at the gastroduodenal junction, in association with quiescence of lhe antrum and proximal duodenum. Manometric studies have also demonstrated localized pyloric pressure waves in humans after ingestion of nutrients or acid, or their intraduodenal infusion (Fisher and Cohen 1973; White et al 1983; Tougas et al 1987; Heddle et al '1988a,b,c & 1989; Houghton et al 1988aab; see also Chapter 5) (Table 4.1 and Figure 4.3). Analysis of IPPWs recorded after intraduodenal infusion of a lipid emulsion (Heddle et al 1988b) indicate that these occur over a narrow zone (Figures 4.4 and 4.5), less than 9 mm wide in 92T" of cases and less than 6 mm wide in 66% (Heddle et al 1988b). Analysis of the timing of pressures recorded in three adjacent closely-spaced side-holes in that study failed to show any consistent pattern of propagation of lPPWs, suggesting they are non-propulsive and likely to retard transpyloric flow (Chapter 6). ln contrast to these sludies, King et al (1984 & 1985) did not observe isolated pyloric contractions after lhe ingestion of dilute orange juice. Because of the lechnical difficulties of ultrasonography (Chapter 2), the very narrow zone of pyloric contraclion, and the short post-prandial sampling times chosen in these studies (approximately 4 minutes per subject), the findings of King and co-workers may not have 38

FASTING DURING LIPID mm Hg 40 Sleeve

0 40 s.H. f 0

0 40 S.H.8

0 40 s.H.6

0 ^^,^.----^-/\,^\-/'\^/ 40 s.H.4

0 ,^-¡=ô--_,r\.\/\-^,^.rr 40 s.H.2

30 seconds

Figure 4.4 Recording from a sleeve/side-hole manometric assembly incorporating 13 side-holes at 3mm intervals along the length of the sleeve, during intraduodenal lipid infusion. Pressure waves recorded by the sleeve and every second side-hole are shown. IPPWs recorded by the sleeve sensor are observed in only one of the side- holes. No pressure rise is recorded by adjacent side-holes, indicating that lumen occlusion is occurring over a very narrow zone. (From Heddle et al 1988b, with permission.)

200

ø rso f,g l! o r00

cE l¡J @ =50z o @ 2 3 1

NUMBER OF SIOE HOLES

Figure 4.5 W¡dth of the phasically active pyloric zone. This frequency histogram shows the number of side-holes, spaced 3mm apart, which recorded 313 consecutive IPPWs stimulated by intraduodenal lipid infusion. lf a pressure wave was recorded by 1 side-hole, its greatest possible width was <6mm, by 2 side-holes

demonstrated the full range of post-prandial pyloric motor activity that occurred in these subjects. lt is also very possible that the calorie content of the meal, which was not stated, may not have been sufficient to stimulate localized pyloric contraction (Chapter 5).

The paltern of IPPWs recognized manometrically could correspond either to fluoroscoPic observations of localized contraction of the gastroduodenal junction, in which antral contraclions of any amplitude were absent (see above), or inductographic reports that localized reductions in pyloric luminal diameter may be associated with only partial reduction of the amplitudes of antral and duodenal contractions (Ehrlein et al 1984). The limitations of antral side-hole manometry (Section 2.g.1.2) are such that the degree of antral suppression associated with IPPWs remains unknown. lf localized phasic pyloric motility is to be established as a likely mechanism for braking gastric emptying (Chapter 6), it is important to demonstrate that the associated suppression of antral pumping includes non-lumen occluding contractions of any amplitude (ChaPter 8)'

Fluoroscopy of the pylorus after ingestion of a meal (Cannon 1898; Cannon and Leib 1911;Shay and Gershon-Cohen 1934; White et al 1983) suggests that sustained pyloric contraction also occurs in the fed state. This belief is supported indirectly by the observations of Tougas et al (1987), demonstrating prolonged pyloric luminal closure after intraduodenal lipid infusion, during the emptying of a liquid barium meal. Recent sleeve/side-hole manometric recordings confirm 'I localized tonic pyloric contraction can indeed occur in the fed state, usually at the same time !l.Ì that i as lPPWs, which are induced by similar stimuli (Heddle et al 1988a&b). lt also occurs over the same narrow zone (Heddle et al 1988b), suggesting that pyloric tone may also reflect contraction of the distal pyloric muscle loop (Chapter 1). Other manometric studies reporting pyloric tone have been less convincing because of unsatisfactory methodology (White et al 1983; Pandolfo et al 1979; Dooley et al 1985; Houghton et al 1988a). With one recent exception, in which the effect of intraduodenal lipid infusion on gastric emptying was studied (Heddle et al 1989), there have been no satisfactory studies of basal pyloric pressure during the emptying of a solid meal'

4.4 CONCLUSION The temporal association of pyloric contractions with those of the antrum and duodenum suggests that pyloric motility is frequently coordinated. Refinements in the rules for defining coordination are needed, and will become possible as understanding of normal motility Patterns increases. The pylorus normally contracls in sequence with the anlrum and duodenum, both during fasting and in the fed state, but localized pyloric motility is also well recognized depending on specific luminal conditions. The patterns of normal pyloric motility in many circumstances remain unknown, however. ln particular, whether pyloric motility changes ilvitn tne rate of gaslric emptying has not been adequately studied. Additional studies are also required to further define the behaviour of the antrum when localized increases in pyloric pressure are recorded. r 40

CHAPTER 5

Control of Pyloric MotilitY

5.1 INTRODUCTION Alterations in the patterns of pyloric contractions (Chapter 4) must ultimately result from ^ ^3 mOOUlatlOn Ol Ing el€lclf lcal COntf Ol aIlO f (JsPuf lljtt ctçtlvlly-^¡:..:¡.. ul Pylvr-.'l^-:^ rv Ðrrrvvrrr^-^^+1.. -',-^l^rrrvovre' Unfortunately, many olherwise exacting electrophysiological studies of this region have failed to 1982). :!f specify the precise structure to which they were directed (for example, Papasova et al information about electrical activity which can be ,] Consequently, there is relatively little attributed confidently to the distal pyloric muscle loop'

Stimulation of small intestinal mucosal receptors, input from the central nervous system, and gastrointestinal hormonos are all likely to be important in regulating pyloric motility. These factors are not mutually exclusive, and it must be recognized that there is considerable overlap in their roles. For example, excitation of duodenal nutrient r€ceptors may stimulate local neural pathways, release of hormones, or both. Furthermore, gastrointestinal hormones may act on the pylorus directly, or via central or peripheral neural pathways.

The following sections will discuss the electrical properties of the smooth muscle at the gastroduodenal junction, with particular attention to the distal pyloric muscle loop. The electrophysiological terminology used throughout this thesis is based on the definitions discussed and used by Sarna (1975 a 1989), which are indicated diagrammatically in Figure 5.1. Possible mechanisms by which pyloric electrical activity may be influenced will then be considered, I I before discussing more specifically the role of the small intestine and the central nervous system I in the regulation of human pyloric motility. For simplicity, the effects of gastrointestinal hormones on pyloric motility are summarized in a section subsequent to this.

! 41

b d -50mV a e

1 ilT

"iiti

Flgure 5.1 Configuration of the potential changes in pyloric smooth muscle cells a) resting membrane potential b) depolarization c) plateau potential, with oscillations and spikes d) repolarization Cyclical, sponlaneous depolarization/repolarizations of lhe membrane, in lhe absence of sustained smooth muscle contraction, are known as the e/ecfrical control activity, and this determines the maximum frequency of smooth muscle contract¡on. The initial rapid depolarization may exceed threshold for smooth muscle contraction, but any resulting contraction is of insignificant amplitude and may not be delected. The plateau potential of the electrical control activity is characteristically smooth and does not reach threshold for contraction or lhe generation of spike potentials. ln contrasl, electrical response activity is characterized by a plateau potential of greater amplilude, often associated with oscillations and spikes, and always associated with a sustained second contraction. (From van der Schee 1974, with permission.)

l 42

5.2 ELECTROMECHANICAL PROPERTIES OF PYLORIC SMOOTH MUSCLE 5.2.1 Characterlstlcs of the Electrical Control and Response Actlvlty of Smooth Muscle Cells at the Distal Pylorlc Muscle Loop Mosl information about the electrophysiological properties of smoolh muscle cells from the distal pyloric ring is derived from in vitro studies, although spontaneous depolarizations have been recorded from the circular smooth muscle cells of the canine pylorus in vivo. This latter finding could not be confirmed by Bass et al (1961), but in this study it is uncertain whether the solid- needle electrodes inserted percutaneously into the gastroduodenal junction were correctly positioned within pyloric smooth muscle. The cellular origin of electrical control activity within the distal pyloric muscle ring remains uncertain. Traditionally, it is thought that electrical activity in the gastrointestinal tract originates from smooth muscle cells within the outer of circular smooth muscle layer (Szurszewski 1987). Recently, this belief has been challenged (Szurszewski 1987; Daniel and Allescher 1990) by studies suggesting lhat eleclrical control aclivity may be generated instead by the interstitial cells of Cajal (Section 1.5).

ln vitro intracellular microelectrode recordings from smooth muscle cells of the distal pyloric muscle ring (Borloff and Davis 1968; Edwards and Rowlands 1968; Fujii 1971 ; El-Sharkawy et al 1978; Szurszewski 1987) have icjentified severai important differences in ihe electrophysiological properties of this region compared to the rest of the stomach (Figure 5.2). The gradient in the frequency of spontaneous electrical control activity, from 0.15/min at the u (El-Sharkawy al I pylorus, 1o 3.7/min in the corpus and S/min in the mid- and orad corpus et 1978), may be one of the most important determinants of the coordination of motility (Sarna 1989). Because of the low intrinsic frequency in the distal stomach, events generated proximally are able to propagate to the pylorus without interference (see below), favouring the occurrence of long, peristaltic contractions (Chapter 4). The smooth muscle cells of the canine distal pyloric ring have the most negative intracellular resting membrane potential of approximately -75 mV, compared with the corpus (-60 mV) and the fundus (-48 mV) (Szurszewski 1987). Proximal to the pylorus, the rate of rise of the upstroke potential and the amplitude of the upstroke and the plateau potentials increase from the corpus to the terminal antrum. Although the amplitude of the upstoke potential at the pylorus is greater than the terminal antrum, its rate of rise is similar, and the amplitude of the plateau potential is actually less (El-Sharkawy et al 1978). The characteristics of human pyloric electrical activity have been less adequately defined.

lmmedialely orad to the proximal pyloric muscle ring there is a transition zone in humans and t dogs where, in contrast to the smooth plateau potentials of antral smooth muscle cells, I oscillalions appear which give rise occasionally to 'spike potentials', which increase in ; frequency towards lhe pylorus where they may occur lhroughout the plateau (Edwards and Rowlands 1968; Papasova 1968; El-Sharkawy et al 1978; Szurszewski 1987) (Figures 5.1 and 5.2). El-Sharkawy et al (1978) found variations in the site of this transition zone, which in one l 43

Fu¡dus .,8 A I -18

Vtry Orod CorPus -2t I I -5t

Orod Corgus -26 c I -56 >-l

ll¡d Orod Corpus o -ro I -60

Covtlod CorPus -23 E I ]--_{ -63 O¡od Anltum F -l9 I -69

O¡od Tcrminol Anltum G -27 -7o

Coudod Torminøl Anlrum H -28 I -7' l--l lO toc Pyloric Rlng o I -75 lO toc

Figure 5-2 lntracellular resting membrane potsnt¡als and spontaneous electrical activity recorded in vitro from different regions of the canine stomach. The highest frequency of the electrical control aclivity is in the orad corpus' ThL configuration of the electrical activity changes towards the distal stomach, and at the pylorus it is characlerized by a more negalive intracellular resting membrane potential, the longest duration and the i occurrence of oscillations and spike potentials throughout the plateau. Note that the mV scale depicted on the right of the figure varies for each site, and the time scale for the pyloric ring is different. (From Szurszewski 1987, with permission.)

I 44

canine preparation appeared to be as far as 6 cm orad to the pylorus. Electrical control activity at the pylorus also has a relatively long duration of about 21-26 sec, compared to 14 sec in the antrum and 12 sec in the corpus (El-Sharkawy et al 1978). Finally, the rate of propagation of electrical control activity increases from the corpus to the terminal antrum. Recordings from the in vivo human stomach suggest that, in the antrum between 1.5 and 4 cm from the pylorus, it is conducted at 0.5 cm/s€c, compared to 2.06 cm/sec'across the pylorus'(Duthie 1971), and 14 cm/sec in the duodenum (Milton, Smith and Armstrong 1955). As a result, the onset of the electrical response activity in the terminal antrum and the pylorus occurs almost simultaneously (Kelly 1980). This possibly represents the electrophysiological basis for the terminal antral contractions descibed by Code (Carlson et al 1976) (Section 4.3.1). ln the intact stomach, electrical control activity at the pylorus does not occur at the intrinsic pyloric rate of 0.1S/min. Rather, it most often appears to be determined by the higher frequency depolarizations of the gastric pacemaker (Carlson et al 1966; Kelly et al 1969; Hinder and Kelly 1977; Papasova et al 1982), located in the mid-corpus in humans, dogs and cats (Alvarez and Mahoney 1922; Goodman et al 1955; Sarna et al 1972; El-Sharkawy et al 1978). Pyloric electrical control (and response) activity has been reported to follow that of the antrum during each phase of fasting myoelectric activity in dogs (Papasova et al 1982), although the position of the electrodes in this study was not clearly indicated. This study also described electric control activity at the pylorus in the rhythm found in the duodenum. Consistent with these observations, regular, localized pyloric contractions (lPPWs, Chapter 4) are often recorded at antral frequency, even in the absence of detectable antral contractions (Heddle et al 1988c a 1989). IPPWs may also occur at a faster frequency of approximately 12lmin (Allescher et al 1988b; Heddle et al 1989), reflecting probable entrainment of the pylorus by the duodenum.

5.2.2 Transmission of Electrlcal Actlvity Across the Pylorus Propagation of electrical activity towards and across the pylorus is likely to be a prerequisite for the coordination of pyloric motility with these adjacent regions (Chapter 4). lnitial evidence suggested that the gastroduodenal junction constituted an impenetrable electrical barrier (Bass et al 1961; Allen et al 1964), however subsequent studies in vivo and in vitro suggest that transmission across the pylorus does occur, and in both directions (Bortoff and Weg 1965; Daniel and Weibe 1966; Bortoff and Davis 1968; Duthie et al 1971 ; Bortoff 1990). Bortoff and Davis (1968) sludied gastroduodenal electrical activity in vitro (see below), and with suction electrodes in vivo. ln cats, rhesus monkeys, baboons and to a lesser extent in dogs, augmentations of proximal duodenal depolarizations were clearly seen to occur in the antral rhythm. These augmentations were small, and were not thought to be of sufficient magnitude to induce electrical response activity, unless additionally inf luenced by neurotransmitters or 45

hormones (Section 5.2.3). Thus, for example, duodenal spike activity was induced by low frequency vagal stimulation only in those slow waves which were augmented in this way.

Transmission of electrical control activity across the pylorus has also been demonstrated in humans (Duthie et al 1971). ln that study, electrical signals were recorded over 3-6 days with subserosal electrodes which had been placed at laparotomy. Electrodes were sited in the antrum and duodenum within 1.5 cm of the pylorus, but no recordings of pyloric electrical signals were made. Stable electrical control potentials were recorded in the antrum al 3.12 +/- 0.05 cycles/min, and in the duodenum 10 lo 12 cm distal to the pylorus at approximalely 12 cycles/min. Electrodes in the proximal 4 to 5 cm of the duodenum recorded fluctuations in a predominantly antral rhythm of 3 cycle/min. These results, however, must be interpreted wilh caution, as recordings were made soon after upper abdominal surgery for pathological conditions.

Antral modulation of duodenal electrical activity has been reported for a variable distance distal to the pylorus. Bortoff and Davis (1968) found an effect only in the proximal 15 to 17 mm of the duodenal cap, although other studies in animals have demonstrated some influence as far as 6-8 cm distal to the pylorus (McOoy and Bass 1963; Cooke and Kotteman 1974). Duthie et al (1971) found an antral rhythm up to 10 cm from the pylorus in humans. A similar influence of duodenal slow waves on antral slow wave depolarization has also been observed (Bortoff and Davis 1968), indicating that electrical transmission may occur across the pylorus in both directions.

Thal transmission of electrical control activity across the pylorus occurs via myogenic pathways is suggested by results of studies demonstrating that it persists after tetrodotoxin in vitro (Fujii 1971) and in vivo (Schuurkes and van Neuten 1984), as well as after anoxic destruction of the myenteric plexus in vitro (Bortoff and Davis 1968). Although it is possible that normal neural mechanisms may be substantially altered or even absenl in vitro, the failure of tetrodoloxin to prevent lransmission in vivo strongly mitigates against a neurogenically mediated effect. Myogenic transmission could occur passively, by interdigitation and electrical coupling of antral and duodenal smooth muscle cell bundles (Torgersen 1942; Bortoff and Weg 1965; Bortoff and Davis 1968), or may involve active generation of electrical activity at either antral or duodenal frequency by pyloric smooth muscle cells (Bortoff 1990). Daniel and Allescher (1990) speculated that interstitial cells of Cajal (Chapter 1) may form a network of coupled cells which may accompany the myenteric plexus as it crosses in continuity from the antrum to the duodenum. Such a network of interstitial cells of Cajal would provide an alternative non- neurogenic mechanism for transmission of electrical activity across the gastroduodenal junction.

Transmission of the electrical control activity within the distal pyloric muscle loop has not been studied. lt is tempting to speculate that the mode of eleclrical transmission may be the same as 46 that in the anlrum, where cellular coupling is rather similar (Chapter 1). Here, it spreads inwards from the cells of origin (Section 5.2.1) lowards the cells next to the submucosal plexus, and then propagates circumferentially and aborally (Kelly et al 1969; Weber and Kohatsu 1970).

5.2.3 Mechanlsms for the Modulatlon of Pyloric Electrlcal Activlty by Neurotransmltters and Gastrolntestlnal Hormones The relationship between electrical activity and contraction of smooth muscle cells from the distal pyloric loop (excitation/contraction coupling), and the ways by which this can be modified, have been inadequately studied. Szurszewski (1987) suggested lhat the amplitude or duration of gastrointestinal smooth muscle contraction would be negligible but for lhe influence of hormones and neurotransmitters. For example, in lhe presence of excitatory stimuli such as acetylcholine or pentagastrin, the voltage of the plateau potential may be raised above the mechanical threshold, leading to a significant increase in the amplitude of the accompanying phasic contraction (Morgan and Szurszewski 1980; Szurszewski 1987). Modification of electrical activity in this way has been reported in studies of smooth muscle from the antrum, fundus and duodenum in humans and animals (Daniel 1965; Bortoff and Davis 1968; Duthie 1971; Morgan et al 1978)' Clearly, similar studies of smooth muscle from the distal pyloric muscle loop are required.

5.2.4 Concluslon The smooth muscle cells of the pylorus are capable of spontaneous depolarization, however, in vivo electrical activity at lhe pylorus is entrained most frequently by the antrum, and sometimes by the duodenum. Electrical activity may be transmitted across the pylorus, probably in both directions. Further, accurate recordings from the distal pyloric muscle loop are required to firmly establish the participation of these mechanisms in the coordination of pyloric with antral and duodenal motility. Modulation of pyloric motor function by neurotransmitters and peptide hormones could occur via changes in electrical response activ¡ty, but specific studies of the pylorus have not yet been performed.

5.3 CENTRAL NERVOUS SYSTEM CONTROL OF PYLORIC MOTOR FUNCTION There is very little direct information about the influence of the cenlral nervous syslem on the pylorus. Recognition of such an influence is important, not least because many in vivo studies of pyloric motility and its control have been performed in anaesthetized animals (see, for example, Allescher et al 1988aab). The brain is known to modify many aspects of gastric function in animals (Pascaud et al 1978; Bueno et al 1983; Konturek et al 1985; Pappas et al 1986) and, recenlly, the centrally acting stimuli cold pain and labyrinthine stimulation have been shown to alter fasting antral and duodenal motility and delay gastric emptying in humans (Thompson et al 1982 & 1983; Stanghellini el al 1983 & 1984). ln none of these studies has pyloric motility been evaluated. Central nervous system effects, however, may be mediated via the vagus or sympathetic nêrves, and good evidence exists that lhe pylorus is influenced by both these routes. 47

5.3.1 Evldence for Control of the Pylorus by the Vagus Nerve 5.3.1.1 Excitatory lnnervation of the Pylorus by lhe Vagus The pylorus is innervated by motor and sensory vagal libres with connections in the brainstem (Chapter 1). Certain intestinal stimuli which are plausible regulators of norr¡al pyloric motility (Section 5.4) have also been reported to influence vagal function. For example, mechanical stimulation of duodenal mucosa by stroking with glass rods (Cottrell and lggo 1984), and intraduodenal infusions of glucose (Mei 1978; Ewarl and Wingale 1984) and hypertonic/hypotonic liquids (Garnier and Mei 1982), stimulate afferent vagal neurones. Furthermore, intraduodenal glucose has been shown to change the firing rates of nuclei in the dorsal motor nucleus of the vagus (Ewart and Wingate 1984). ln anaesthetized cats, Pagani et al (1985) found that electrical stimulation of the dorsal motor nucleus of the vagus caused an alteration of pyloric motility, but it is uncertain whether the slrain gauges in that study were sutured over the distal pyloric muscle loop. Sleeve manometry indicates that central vagal stimulation by intracerebro-ventricular injection of a lhyrolropin releasing hormone analogue induces tonic and phasic pyloric contraction in anaeslhetized rats (Raybould et al 1989a). Electrical stimulation of the ends of sectioned vagi also indicales the presence of an excitatory molor innervalion of the pylorus by the vagus (Langley 1898; Edin et al 1979a & 1980; Mir et al 1979; Allescher et al 1988b). Most convincingly, Allescher et al (1988b) recorded pyloric motilily with a sleeve sensor in anaeslhetized dogs, and found that low frequency stimulation (0.1-0.5 Hz) of the distal vagal stump induced localized phasic pyloric contraclions. This vagal excitalory response was abolished in dogs by atropine, hexamelhonium and tetrodotoxin (Allescher et al 1988b), and in cats by tetrodotoxin, phentolamine and a combination of atropine and hexamelhonium (Mir et al 1979; Telford et al 1979). Consistent wilh these observalions, Edin (1979a a Edin et al 1980) found that efferent vagal stimulation reduced transpyloric flow in anaesthetized cals, but the techniques used in this study were unsuilable (Chapter 2). Reduced transpyloric flow (Edin 1980) and contraction of the gastroduodenal junction in cats (Behar et al 1979) have been reported after stimulation of the central end of one vagus, with the contralateral vagus left intact, suggesting the presence of an excitatory vago- vagal reflex. These latter stud¡os are also difficult to interpret and require confirmation with adequate methodology. Despite these findings, the effects of vagotomy on pyloric motility in animals have been inconsistent (Ludwick 1969; Roman 1982; Aeberhard 1977; also see below).

5.3.1.2 lnhibitory lnnervalion of lhe Pylorus by the Vagus Higher frequency vagal stimulation (0.7-50 Hz) inhibits the pylorus in anaesthetized dogs and cats (Behar et al 1979; Mir et al 1979; Allescher et al 1988b) via a non-cholinergic, nonadrenergic neural pathway (Anuras 1974; Mir 1979; Allescher et al 1988b), consistent with the findings of Deloof and Rousseau (1985) in vagotomized rabbits. This vagal mechanism was able to 48

suppress pyloric excitation produced by duodenal field stimulation (Allescher et al 1988b), and may correspond to the inhibition of pyloric smooth muscle strips by electrical field stimulation (Anuras et al 1974). The same electrical slimulus induces anlral and duodenal contractions after a brief latency (Mir et al 1978; Behar et al 1979; Allescher et al 1988b). Such reciprocal innervation could be important in the production of localized pyloric contraction (Chapler 4).

Vagotomy in rabbits was reported 1o cause increased electrical response activity which was not affected by alropine, phentolamine or propranolol, also suggesting a non-cholinergic, nonadrenergic vagal inhibitory innervation of the pylorus (Deloof and Rousseau 1985). Consistent with this, Riddell et al (1989), using more rigorous methodology, found that vagotomy in pigs led to unrem¡tting localized pyloric pressure waves. ln contrast, vagotomy had no effect on basal pyloric pressure in cals (Reynolds et al 1985) and no consistent effect on spontaneous or acid- induced pyloric contraction in dogs (Allescher et al 1988b).

ln conclusion, animal studies have convincingly demonstrated pyloric innervation by both inhibitory and excitatory vagal fibres. There is no information about vagal modulation of pyloric motility in humans.

5.3.2 Sympathetlc Control of Pylorlc Motllity The distal pyloric muscle loop contains a dense adrenergic innervation (Furness and Costa 1987; Allescher et al 1989). Detailed stud¡es with strain gauges and sleeve/side-hole manometry (Allescher et al 1989) indicate the presence of neural p receptors which can excite pyloric smooth muscle via acelylcholine release, or inhibit it by acting on other p receptors. lnhibitory 6¡1 and cr,2 receptors were also found on neurones to the pylorus. This complicated innervation, together with the dilficulties of recording from the distal pyloric muscle loop, may explain the apparent conflicts among previous studies of the sympathetic control of pyloric motor function. Deloof (1988) found evidence of inhibition of pyloric electrical spike activity in rabbits with the adrenergic agonists phenylephrine, clonidine and salbutamol. ln addition, he reported that electrical stimulation of the peripheral end of the thoracic sympathetic trunk in decorticate rabbits generally inhibited spontaneous and vagally-stimulated pyloric spike bursts, a response abolished by p adrenergic blockade, but not by phentolamine. Stimulation of the central end of one sympathetic trunk also inhibited pyloric spike bursts, an effect abolished by contralateral sectioning and phentolamine. These manoeuvres also inhibited antral activity, consistent with the previously reported sympathetic inhibitory innervation of non-sphincteric gastrointestinal smoolh muscle (Furness and Costa 1987). The findings of Biancani et al (1981) supporl the existence of inhibitory sympathetic innervation of the pylorus, since the excitalory response of the pylorus to histamine was increased by a adrenergic blockade and inhibited by an o agonist. 49

Othor studies, however, suggest the existence of an excitatory sympathetic innervation. Splanchnic nerve slimulation in rabbits has been reported to increase the frequency of pyloric electrical rosponse activity in rabbits (Deloof 1988). Lerman et al (1981) reported an excitatory cholinergic sympathetic pathway in dogs, and infusion of noradrenergic agonists was found to stimulate tonic contraction of the rat pylorus (Scheurer et al 1983). Neuropeptides other than those mentioned above may also be involved in sympathetic innervation of the pylorus. For example, neurons projecting to the region of the "pyloric sphincter' from coeliac and dorsal root ganglia (T7-10) in the rat stain for calcitonin gene-related peptide, subslance P and neuropeptide Y (Plotkin and Reynolds 1989) (see also Section 5.5). The physiological role of these peptides remains unknown. ln humans, Stanghellini st al (1983) demonstrated that the delay in gastric emptying induced by cold pain and labyrinthine stimulation was associated with elevations in plasma levels of nor-epinephrine and p-endorphin, and were inhibited by p-adrenergic receptor blockade. There have no direct studies of the role of sympathetic nerves in the control of human pyloric motility.

ln conclusion, there is evidence for both excitatory and inhibitory pyloric innervation by sympathetic libres, and very little consistent informalion on the effect of sympathetic stimulation on pyloric motility. Further studies are needed in suitable animal models and in humans with adequate manometric techniques (see Chapter 2).

5.4 CONTROL OF THE PYLORUS BY THE SMALL INTESTINE ln humans and animals, localized pyloric contraction (Chapter 4) has been demonstrated in response to the presênce in the small intestine of a variety of intraluminal stimuli. This stimulation of pyloric motility is likely to be due lo excitation of small intestinal mucosal receptors. Although there is some information from animal studies, little is known about the mechanisms mediating this effect in humans, although both neural pathways (Section 5.4.3) and the release of hormones (Section 5.5) are likely to be involved. This section discusses the control of pyloric motility by the stimulation of small intestinal mucosal receplors, and the possible mechanisms mediating this control.

5.4.1 The Effect of lntraluminal Small lntestlnal Stimull on Pylorlc Motility 5.4.1.1 lnlraluminal Nutrients There is unequivocal evidence from many studies, using a variety of techniques, that the presence of nutrients in the upper small intestine influences pyloric motility. lngestion of nutrient meals or intragastric infusions of liquids containing fat or carbohydrate, are associated with increased phasic pyloric contraction (White et al 1983; Houghton et al 1988b) and a reduced diameter of the pyloric lumen (Kaye and Ehrlein 1983). Localized phasic and tonic pyloric motility or reductions in pyloric luminal diameter have been reported in animals in response to 50

intraduodenal fat (Keinke et al 1983; Ehrlein et al 1984) and glucose (Erhlein et al 1984). Reynolds et al (1985) found that intraduodenal infusions of tryptophan or a mixed amino acids induced tonic pyloric contraction in cats, a response blocked by naloxone (see below). lnterestingly, no such stimulation was associated with phenylalanine, suggesting that this effect on the pylorus may be specific to certain amino acids. ln humans, the sleeve manometric studies of Heddle and co-workers (Heddle et al 1988b4c, 1989) and Bovell et al (1987) clearly demonstrale an increase in the number of isolated pyloric pressure waves (lPPWs) and a rise in basal pyloric pressure (Chapter 4) afler intraduodenal inf usions of lntralipid, a hypertonic hydrolysed fat emulsion, and hypertonic carbohydrale. Several other less satisfactory studies have reported similar results (Fisher and Cohen 1973; Fisher et al 1973). The rosponse to dextrose is dose dependent, and maximal at an infusion rale similar to the known emptying rate of dextrose from the stomach (2.1 Kcal/min) (Heddle et al 1988c). Although the pyloric motor response to dextrose occurs within 5 minutes, the effects of lipid are often observed after a latency of up to 15 to 20 minutes (Best 191'l cited by Thomas 1957; Heddle et al 1988b & 1989). This difference probably reflects the time taken for partial digestion of lipid, a requirement for the delay in gastric emptying also produced by this stimulus (Cortot et al 1982). Recently, Fraser et al (1989b) have shown that the phasic pyloric motor rssponse is sustained for at least 90 minutes during continuous intraduodenal lipid infusions.

There have been no studies of the effect of nutrient stimulation of the distal small intestine on pyloric motility. lt is probable, because of lhe retardant effect of ileal nutrient infusions on gastric emptying, and the likely mechanical role of the pylorus (Chapter 6), that the distal small bowel also participates in the control of pyloric motility. This requires specific investigation.

5.4.1.2 Small lntestinal Acidificalion Gastric emptying was once believed to be primarily controlled by the effects of acid on pyloric motility, with intragastric acid causing pyloric relaxalion, and intraduodenal acid inducing pyloric contraction (Cannon 1907 & 1911;Best 1911 cited by Thomas 1957). Pyloric closure in response to acid was described by Pavlov (1901, see Thomas 1957). Recently, duodenal acidification was shown to increase feline pyloric electrical response activity and to be associated with phasic and/or tonic pyloric contraction in dogs and cats (Allescher et al 1988b&1989c; Reynolds et al 1984). Sleeve/side-hole manometry has convincingly demonstrated a phasic and tonic pyloric motor response to duodenal acid in humans (Houghton et al 1987; Tougas et al 1988a).

5.4.2 Slte and Nature of Mucosal Receptors ln the Small Intestlne The pyloric motor responses to the intraluminal stimuli listed above are likely to be mediated by receptors located on or near the small intestinal mucosa. For example, lhe effect of duodenal 51

acidification on phasic pyloric contraction in cats and dogs is prevented by the luminal application of the local anaesthetics ethyl aminobenzoate and lignocaine respectively (Reynolds et al 1984; Allescher et al 1987). Recent studies suggest that some mucosal receptors may be evenly distributed throughout the small bowel, while others may be quite localized. Cooke (1977) found that perfusion of an isolated segment of jejunum, but nol duodenum, delayed gastric emptying in dogs. Lin et al (1989aab) showed, in dogs with small bowel fistulae at different levels, that receptors for glucose were present in both the proximal and distal small intestine, whereas those for acid appeared to be confined largely lo the proximal gut, especially lhe duodenum (Cooke 19741. The intensily of the inhibition of gastric emptying produced by intraduodenal adminislrat- ion of glucose or acid has been shown to be dependent on the length of mucosa, and hence the number of mucosal receptors, exposed to these slimuli (Cooke 1974; Lin st al 1989aab). The effect of regional intestinal stimulation on pyloric motility has not been examined.

It is unclear whether these receptors respond to calories, osmolarity, or both. Studies have been inconclusive on this point, largely because most nutrient-rich solutions shown to stimulate the pylorus have also been of high osmolarity. Although some investigators have discounted the role of specific nutrient receptors (Cooke 1977 & 1978; Parr et al 1987), recent information suggests that nutrients may be a more potent stimulus of pyloric motility than equimolar non- nutrients (Heddle 1988c). The observations of Lin et al, noted above, argue for different stimuli having separate receptors, so the relative importance of calories and osmolarity deserves further study.

It has also been suggested that the pylorus may interact dynamically with the antrum and duodenum via mechanoreceptors, shown to be located near the pylorus and in the proximal and distal small bowel in cats, sheep and humans (Paintal 1954; Mei 1970; Bitar et al 1975; Cottrell and lggo 1984; Rouillon et al 1989). Mucosal mechanorecoptors in animals are responsive to distension by intraluminal balloons, mucosal stroking with glass probes or cotton wool, and increases in muscle tension related to spontaneous or induced duodenal contractions (Cottrell and lggo 1984). A role for mechanoreceptors in controlling the human pylorus remains entirely speculative.

5.4.3 Candldate Local Neural Mechanisms for Mediating the Small lntestinal Control of Pyloric Motility There is considerable specialization of neural slructures at lhe level of the distal pyloric muscle loop (Chapter 1). Electron microscopy in the guinea pig (Cai and Gabella 1984) and lighl microscopy in the dog (Daniel and Allescher 1990) have shown that the distal pyloric muscle loop contains a distribution of neuropeptides which differs significantly from that in the adjacent pyloric segment and antrum. ln vitro studies of field-slimulated muscle strips suggest, despite some concerns about sampling errors (Chapter 2), the presence of a non-adrenergic inhibitory 52

innervation which is not demonstrable in the adjacent muscle of the antrum or duodenum (Anuras et al 1984). Recent studies by Allescher and co-workers have yielded considerable information about the neurohumoral control of the canine pylorus (Daniel and Allescher 1990). This section describes the intrinsic neural pathways and neurotransmitters likely to be involved in mediating the effect on pyloric motility of lhe intestinal stimuli discussed in Section 5.4.1.

5.4.3.1 Neural Palhways Connecting the Duodenum 1o the Pylorus The pyloric motor response lo duodenal nutrient/osmotic mucosal receptor stimulation may be mediated by local nsural pathways. Thus, for example, pyloric contraction has been induced by intraduodenal infusions of phenylbiguanide, a specific sensory nerve stimulant, and the local anaesthetic lignocaine reduces the pyloric motor rosponse to intraduodenal acid (Allescher et al 1987). Furthermore, duodenal field stimulation 3-5 cm aborad to lhe pylorus induces phasic and tonic pyloric contractions in the dog (Allescher et al 1988b) which are blocked by atropine, hexamethonium, letrodotoxin and duodenal transection, but not vagotomy. Pyloric pressure waves slimulated by intraduodenal dexlrose in pigs are also reduced by duodenal transeclion (Treacy et al 1988b). These findings indicate the existence of an intramural, orally projecting, cholinergic, multineuronal pathway connecting duodenal receptors to the pylorus. Consistent wilh these studies, Valenzuela et al (1976) reported that intraduodenal infusions of acid in humans caused an increase in pyloric pressure which was abolished by atropine, but their manometric technique did not permit reliable pyloric measurements (Chapter 2).

Opiate receptors may be involved in the pyloric motor responses to duodenal slimulation under some circumstances. lntra-arterial injections of leu- and met-enkephalin, as well as duodenal acidification, cause increases in localized pyloric contraction in cats which are blocked by tetrodotoxin and naloxone (Reynolds et al 1984 & 1985). ln humans, inlravenous infusion of p endorphin may increase pyloric pressure (Camilleri et al 1986), but naloxone had no effect on lipid-induced localized pyloric contraction (Bovell et al 1987). A report that histamine stimulates pyloric contractions in cats (Biancani et al 1981)is unconfirmed, but histamine does not seem lo mediate the feline pyloric motor response to intraduodenal acidification, since this response is not blocked by cimetidine (Reynolds et al 1984). Additional studies, especially to evaluate the eflect of specific neurotransmitter antagonists, are needed to further elucidate the pharmacology of the pyloric motor responses to intraduodenal nutrients and other stimuli in humans.

5.4.3.2 Pathways Connecting the Antrum to the Pylorus Little is known about the influence of antral mucosal receptor stimulation on pyloric motility. Electrical stimulation of intramural antral nerves potently inhibits phasic and tonic pyloric motility in dogs, regardless of whether the contractile activity is spontaneous or induced by field stimulation of duodenal nerves (Allescher et al 1988b). The neurotransmitters mediating this 53

responsê ars unknown. Physiological stimuli which produce this inhibitory effect have not been idenlified, but antral distension by intragastric contents is a candidate.

5.4.3.3 Conclusion The pylorus has a specialized complement of neurotransmitters and is innervated by intramural nerves from the duodenum and antrum. These local neural pathways are likely to participate in mediating small intestinal nutrient/osmotic receptor feedback to the pylorus. Cholinergic mschanisms appear to be parlicularly important in animals, but the pharmacology of pyloric motor control has not been properly evaluated in humans.

5.5 CONTROL OF THE PYLORUS BY GASTROINTESTINAL HORMONES Numerous hormones secreted by the gastrointestinal tract have been implicated in the regulation of gastrointestinal motility and gastric emptying, but little is known about their effects on the pylorus. The circular smooth muscle of the normal pylorus is richly supplied with nerve fibres reactive for enkephalin, neuropeptide Y, substance P and vasoactive intestinal polypeptide (Wattchow et al 1987 & 1988; Daniel and Allescher 1990), howsver the function of lhese peptides has not been elucidated. A number of gastrointestinal hormones or their receptors are also found within the central nervous system and/or in vagal nerue fibres (Section 5.3).

Criteria for establishing a physiological role for a hormone wers proposed by Grossman (1974). To be implicated in mediating the pyloric motor rêsponse to a specific slimulus, a rise in plasma levels of the hormone should be demonstrated in association with stimulus and response, exogenously administered hormone should reproduce a similar motor response at doses which are judged to be physiological, and the physiological response should be blocked by specific hormone antagonists. No candidate hormone has yet fulfilled all these criteria for the control of pyloric motility. lnterspecies variations, and differences in hormonal effects on smooth muscle from different sites, even within the distal pyloric muscle loop itself (Golenhofen et al 1980), must be taken into account. Hormone preparations may contain impurities, assays are often technically difficult, and there are problems in determining the true physiological activity of hormone sub- fraclions, such as cholecystokinin octapeptide (CCK-8) (personal communication). Other difficulties in demonstrating a role for hormones in the control of the pylorus relate to lhe demands of in vivo and in vitro measurements of pyloric motility (Chapter 2). The following section reviews the literature on those gastrointestinal hormones which are likely to be important in the control of pyloric motility in humans.

5.5.1 Cholecystoklnln Cholecystokinin (CCK) is a peptide hormone which exists naturally in a number of molecular forms, the most common being the 33 residue CCK-33 and the octapeptide fragment CCK-8. CCK 54

is secreted by endocrine cells in the duodenum and jejunum (Walsh 1987), but is also present in other tissues including the brain (Dockray 1988). CCK receptors are present in lhe circular smooth muscle of the pylorus in rals and monkeys (McHugh and Moran 1986).

Probably because of the experimental difficulties mentioned above, in vitro studies of the effect of CCK on the pyloric smooth muscle have yielded inconsistent results. While CCK was reported to increase the tension in muscle strips from the gastroduodenal junction in opossums (Anuras et al 1974; Lipshutz and Cohen 1972; Fisher et al 1973) and Tucker rats (Berk et al 1986), no effect could be demonstrated on smoolh muscle strips taken from the distal pyloric muscle ring in humans or dogs, at doses of CCK thought to be physiological (Golenhofen et al 1980; Ludtke et al

1 e83). ln vivo, Reynolds et al (1985) demonstrated a dose dependent increase in electrical spiking activity at the gastroduodenal junction after intravenous CCK-8 in cats. Plasma levels of CCK were nol measured in this study, and the electrodes may not have been precisely sited in the distal pyloric muscle loop. CCK has been reported to increase basal pyloric pressure and phasic pyloric contractions in rats with doses judged to be physiological (Scheurer et al 1983; Raybould et al 1989a), and to increase the number or amplitude of phasic pyloric contractions in dogs (Reynolds et al 1985; Telford et al 1979). ln dogs, this stimulatory effect of CCK was blocked by the specific CCK antagonist CR-1392 (Allescher et al 1988c). Neither the pyloric motor response to duodenal acidification in dogs (Allescher et al 1988c), nor the delay in gastric emptying induced by duodenal acidification or glucose infusions in rats (Raybould 1989), are blocked by CR-1392 or the antagonist L364,718 respectively, suggesting that CCK may not mediate these part¡cular effects. ln contrast, the delay in gastric emptying induced by intraduodenal protein and soya bean trypsin inhibitor in anaesthetized rats is inhibited by both L364,718 and perivagal application of the sensory neurotoxin capsaicin, consistent with these effects being mediated by an action of CCK on vagal sensory pathways (Raybould 1989). ln humans, plasma concentrations of CCK rise from a fasting level of 1 pmol to 6 pmol after ingestion of a mixed meal (Liddle 1986). CCK is released by intraduodenal fat, protein, amino acids and possibly also by carbohydrate (Liddle et al 1985 & 1986; Tougas et al 1988b) and duodenal acidification (Barbezat and Grossman 1979), all of which stimulate localized pyloric contraction and delay gastric emplying. lntravenous infusion of exogenous CCK-8, in doses judged to be physiological, slows gastric emptying in many species, including man, rhesus monkeys and dogs (Debas et al 1975; Yamagishi and Debas 1978; Chey et al 1970; Moran and McHugh 1982; Liddle et al 1986; Kleibeuker et al 1988a; Meyer et al 1989). The delay of gastric emptying of normal saline produced by CCK-8 is inhibited by the specific CCK antagonist loxiglumide (Fried et al 1989). Furthermore, loxiglumide significantly increases the rate of gastric emptying of mixed 55

nutrient and zOVo (but not 5%) glucose meals in humans (Fried et al 1989; Meyer et al 1989; Werth et al 1989). Pyloric motility was not measured in these studies, but recently it has been shown that exogenous CCK-8 induces phasic pyloric pressure waves in humans in a dose- dependent fashion, and increases basal pyloric pressure (Fraser et al 1989a). lt is not clear, however, whether these latter effects were achieved at physiological concentrations, since plasma levels were not measured in that study. The increased basal pyloric pressure reported by Fisher et al (1973) and Phaosawasdi and Fisher (1982) was recorded by side-hole pull-throughs and must be interpreted with caution (Chapter 2).

CCK may exert its effect on smooth muscle partly via stimulation of neural receptors. The excitatory effect of CCK on pyloric smooth muscle in rats is due both to a direct action and to an indirect action secondary to release of neurotransmitters, possibly noradrenaline f rom sympathetic nerve endings (Scheurer et al 1983) or acetylcholine (Raybould et al 1989a). Vagal afferents may be required for the retardant effect of CCK on gastric emptying (Dockray 1988), but do not seem to mediate the motor effects of CCK on lhe rat pylorus (Raybould et al 1989a).

5.5.2 Secretln Secretin is a potent pancrealic secretagogue and is released by intraduodenal acid (Schaffalitzky de Muckadell 1978) and possibly by fat (Faichney et al 1981), stimuli which increase localized pyloric contraction and delay gastric emptying (see above). lntravenous infusions of secretin, at doses thought to be physiological, delay gastric emptying (Chey et al 1970; Valenzuela and Deflippi 1981; Kleibeuker et al 1988b). Non-physiological doses of secretin increase the tension in smooth muscle strips from the opossum gastroduodenal junction (Lipshutz and Cohen 1972; Fisher et al 1973), and may stimulate phasic and tonic pyloric contraction in sheep (Ruckebusch and Malbert 1986). lncreases in basal pyloric pressure in humans have only been reported at serum levels of secretin which exceed those found after duodenal acidification, in studies which may also be considered technically inadequate (Fisher et al 1973; Phaosawasdi and Fisher 1982).

5.5.3 Peptide YY Peptide YY (PYY) is a 36 amino acid peptide hormone, similar to pancreatic polypeptide and neuropeptide Y. As yel, there have been no studies of the effect of PYY on the pylorus, but interest in this hormone is justified by its other reported effects. For example, PYY is released by infusion of fat into lhe distal ileum, where the hormone is found in mucosal endocrine cells in its greatest concentralion (Taylor 1985; Tatemoto et al 1988), as well as after a fatty meal, a known stimulus of pyloric contraction (Chapter 4). Both these stimuli delay gastric emptying (Read et al 1982; Spiller et al 1983; Houghton et al 1988b). When infused intravenously, PYY also retards gastric emptying, as well as inhibiting gastric acid secretion (Allen et al 1984; Adrian et al 1985; Pappas et al 1986; Savage et al 1987). 56

5.5.4 Other Candldate Hormones The canine pylorus contains a high density of vasoactive intestinal polypeptide (VlP) and galanin immunorsactive fibres (Alumets et al 1979; Gonda et al 1990). Allescher et al (1989a) demonstrated in dogs that intra-arterial VIP inhibited the phasic and tonic pyloric contractions induced by duodenal field stimulation and acidification. A similar effect was found lor galanin and peptide histamine precursor (PHl). The effects of VIP and PHI were insensitive to tetrodotoxin, suggesting a direct action on smooth muscle. VIP is also found in vagal fibres (Lundberg et al 1979), and both VIP and PHI are candidate transmitters for the well documented non-adrenergic, non-cholinergic inhibitory innervation of the pylorus (Anuras et al 1974; Mir et al 1979; Deloof and Rousseau 1984; Allescher et al 1988b). Sleeve and strain gauge recordings in anaesthetized dogs have shown that intra-arterial infusions of a variety of other neuropeptides, such as substance P, neurokinin A and neurokinin B, also stimulate the pylorus (Allescher et al 1988c).

A number of other hormones, such as gastric inhibitory polypeptide (GlP) (Brown et al 1975), glucagon (Phaosawasdi and Fisher 1982) and neurotensin could be involved in the control of pyloric motility. GlP, for example, is released promptly into plasma after intraduodenal infusions of dextrose (McOullough et al 1983), and intravenous infusion of GIP slows gastric emptying. Glucagon is released by duodenal acidification and hypoglycaemia, slows gastric emPtying and has been reported lo increase basal pyloric pressure at very high concenlrations (Phaosawasdi and Fisher 1982). The physiological effects of these hormones on the pylorus remain unknown.

Several hormones found predominantly in the central nervous system could also be important. Corticotropin releasing factor (CRF) is present in the hypothalamus in a variety of species (Pappas et al 1986) and may be involved in the integrated response to stress in rats (Brown et al 1985; see Chapter 11). Although CRF inhibits lhe interdigestive migrating motor cycle in dogs (Konturek et al 1985) and delays gastric emptying in animals (Pappas et al 1987), its effects on the pylorus have not been studied. Thyrotropin releasing hormone, however, has been reported to induce phasic pyloric contractions, via central vagal stimulation in anaesthetized rats (Raybould et al 1989a). Central administration of calcitonin and calcitonin gene-related peptide interrupts fed patterns of gastric motility in rats (Bueno et al 1983) and potently inhibits gastric emptying, possibly via an adrenergic increase in pyloric motility (Kurosawa et al 1988; Raybould et al 1989b). Calcitonin has also been reported to inhibit pyloric motility in rats via vagal Pathways (Kurosawa et al 1988), but this requires confirmation with rigorous manometry.

5.5.5 Summary Gastrointestinal hormones, in particular CCK, are likely to play a significant role in the control of human pyloric motility. Further testing with physiological doses of exogenous hormones, suitable antagonists and appropriate manometric techniques is required. 57

5.6 CONCLUSTON Although the pylorus is capable of generating its own intrinsic electrical activity, the frequency of pyloric electrical control activity in vivo is usually determined by that in the anlrum, or less commonly the duodenum. Theoretically, pyloric electrical activity could also be influenced by neurotransmitters and gastrointestinal hormones. The pylorus is supplied by vagal and sympathetic fibres, but the role of these neural pathways or the central nervous system is unknown. Whether the central nervous system has any influence on the human pylorus could initially be addressed by measuring the effect on pyloric motility of centrally-acting stimuli, such as labyrinthine stimulation or cold stress. The intrinsic neural innervation of the pylorus has been well investigated in the dog, in which ascending intramural cholinergic neural pathways are especially important in mediating the effects of upper small intestinal nutrient/osmolic receptor stimulation. The neurotransmitters and hormones involved in the regulation of human pyloric motility remain inadequately studied, but there is substantial evidence that cholecystokinin is likely to be physiologically important. 58

CHAPTER 6

The Mechanical Role of the Pylorus

6.1 INTRODUCTION Early this century, it was generally accepted that the pylorus was important in controlling the rate of gastric emptying (for example, Cannon 1898 & 1907; Shay and Gershon-Cohen 1934). This view arose partly because the gross anatomical features of the pylorus (Chapter 1) suggested that it should behave as a sphincter, but predominantly because of the findings and persuasive arguments of Cannon. Cannon obserued pyloric motility fluoroscopically in cats during the emptying of semi-solid meals. He believed that the pylorus was an important regulator of gaslric emplying, relaxing in response to acid in the antrum, and contracting in response to acid in the duodenum (Cannon 1907; Section 5.4.3). This view fell into disrepute because of the inability of later investigators to conf irm a zone of high pressure at the level of the gastroduodenal junction (Wheelon and Thomas 1922; Atkinson el al 1957; Thomas 1957; Louckes et al 1960; Andersson and Grossman 1965; Carlson et al 1966). Thus, Thomas staled in 1957 that 'the most persistent and yet the most thoroughly disproven of all the errors with respect to the pylorus is the idea that the pyloric sphincter is the primary agent in regulating gastric emptying'. The close association of antral, pyloric and duodenal electrical activity reported by Bortoff and Davis (1968; Chapter 5) appeared to supporl the evidence from motility recordings suggesting that the pylorus merely functioned as if it were part of the terminal antrum. This concept was a serious underestimation of the mechanical role of the pylorus, and resulted from inadequate recognition of the methodological challenges involved in accurate pyloric manometry. Recently, the application of suitable recording techniques has enabled the mechanical role of the pylorus to be effectively evaluated for the first time. 59

With few exceptions, intraluminal stimuli which induce localized pyloric contraction (Chapter 5) have been demonstrated in other studies to be associated with delayed gastric smptying (Hunt et al 1956; Thomas 1957; Elias 1968; Hunt and Knox 1968; Cooke 1975 & 1978; Hunt and Stubbs 1975; Brener et al 1983; Keinke and Ehrlein 1983; Houghton et al 1988b). lf the pylorus is important in controlling lranspyloric llow, it should therefore be possible to show that changes in localized pyloric motility correlate with concurrently measured alterations in the rate of gastric emptying. There have been few such studies in humans. Furthermore, if judgements are then to be made about the significance of observed changes in pyloric motility, other mechanisms which participate in the regulation gastric emptying, including fundal tone (Azpiroz and Malagelada 1985aab), antral pumping (Keinke and Ehrlein 1983; Camilleri et al 1985) and duodenal resislance (Miller et al 1981; Keinke and Ehrlein 1983; White et al 1983) should at least be measured and, if possible, controlled for. ln animals, however, the elimination of gastric variables by fundal barostal, antrectomy or duodenectomy produces an unphysiological model, and an appropriate balance between these conflicting issues may be difficult to achieve. ln humans, concurrent measurements of antral, pyloric and duodenal motility and gastric emptying are feasible (Chapters 2 and 3), but measuring fundal lone at the same time presents a major challenge and, with the exception of fundal tone, controlling any of these factors is not possible with presently available techniques. ln this chapter, the evidence that localized contraction of lhe pylorus has a significant effect on transpyloric flow will be reviewed. Emphasis will be placed on the importance of the pylorus in the control of gastric emplying, but the possible role of lhe pylorus in the prevention of retrograde flow from the duodenum to the antrum (duodenogastric reflux) will also be mentioned.

6.2 THE PYLORUS AS A GASTROINTESTINAL SPHINCTEH Demonstration that the pylorus has the functional characteristics of a gastrointestinal sphincter would strengthen the argument that it has a mechanical effect on transpyloric flow. Gastro- intestinal sphincters usually manifest particular properties. For example, suitable stimulation proximal to the sphincter should cause relaxation and distal stimulation should induce contraction with an increase in sphincter pressure (Papasova 1989). Furthermore, in the absence of peristaltic contractions, intraluminal pressure within the sphincter zone should be higher than in adjacent non-sphincteric regions. There is now substantial evidence that the pyloric sphincter may satisfy all three of these conditions. Thus, allhough usually producing no high pressure zone in the unstimulated state, the pylorus is undoubtedly capable of localized tonic contraction under specific intraluminal conditions (Chapter 4). Localized contraction of the pylorus is induced by stimuli acting distal to it, for example via excitation of small intestinal mucosal receptors (Chapter 5). Furthermore, proximal stimuli, such as electrical stimulation of intramural antral nerves, inhibit pyloric contraction (Chapter 5). Though the pylorus can therefore exhibit the 60

properties of a sphincter, it is nevertheless atypical since it does not always manifest a high pressure zone, and in addition should be regarded as highly specialized in view of its anatomy and complex repertoire of coordinated motor Patterns (Chapters 1 and 4).

6.3 THE MECHANICAL EFFECT OF LOCALIZED PYLOR¡C CONTRACTION The lack of any consistenl propagation of localized pyloric contractions (Section 4.3.2) suggests that these events are likely to be non-propulsive, and lhere is experimenlal evidence that localized pyloric contraction does indeed have a retardant effect on the transpyloric flow of intraluminal contents. Approaches for investigating this issue have usually involved measuring alterations in the rate of gastric emptying after surgical removal of the pylorus or interference with ils ability to occlude the lumen. Few studies have correlated accurate recordings of pyloric motility with gastric emptying, although such studies are likely to provide the most information about the mechanical role of the pylorus in humans. The evidence that localized contractions of the pylorus participate in the control of transpyloric flow will be reviewed in this section. Since the patterns of gastric emptying of digestible solids, non-digestible solids, and liquids are different (Kelly 1980; Meyer 1987), and are probably regulated by different motor mechanisms, the function of the pylorus in these processes wil! ba discussed separately.

6.3.1 The Pylorus and Llquld Gastrlc Emptylng Studies of canine pyloric and duodenal motility and intraluminal fluid movements, using serosal slrain gauges, induction coils and fluoroscopy during the emptying of meals of varying viscosity, failed to demonstrate any independent role for the pylorus for non-nutrient and low-nutrient meals (Prove and Ehrlein 1982). Antral contract¡ons, whose amplitude was responsive to the viscosity of the meal, were thought to be the main determinant of gastric emptying. These results were supported by studies of the emptying of non-nutrient isotonic saline in humans (White et al 1981). Under these conditions, however, the pylorus probably functions as part of the antral pump, participating in coordinated antroduodenal waves (Chapter 4). The pylorus is more likely to play an independent role in the emptying of nutrient-rich, acidic or hypertonic liquids (Section 5.4), which proceeds more slowly than the volume-dependent emptying of non- nutrients (Hunt and Spurrell 1951;Hopkins 1966; Elias et al 1968; Hunt and Knox 1969; Collins et al 1983; Houghton et al 1988a). Consistent with this, Ehrlein (1988) reported that flow of chyme into the duodenum occurred when the pylorus was at its widest diameter.

Pyloric excision in animals has been shown to result in increased rates of emptying oÍ 5!" dextrose as well as non-nutrient liquids (Hinder and San-Garde 1983; Muller-Lissner and Blum 19S4). Surgical destruction of the pylorus by pyloroplasty has lead to more rapid (Yamagishi and Debas 1978; Hinder and Bremner 1978; Muller-Lissner and Blum 1984), less rapid (Ormsbee and Bass 1976) or unchanged rales of liquid gaslric emptying (Meyer et al 1979; Kondo et al 1985). 61

The increased rate of emptying in one study (Muller-Lissner and Blum 1984) was negated by an equivalent increase in duodenogastric reflux, but this nevertheless supports a rstardant effect of the pylorus on transpyloric flow in either direction. Pyloroplasty has also been shown to abolish the increased resistance to gastroduodenal flow induced by intraduodenal fat and acid infusions (Miller et al 1981). Attempts at selectively removing the influence of the pylorus by transpyloric stenting (Crider and Thomas 1937; Stemper and Cooke 1976; Kondo et al 1985), on lhe other hand, have not demonstrated any acceleration in the rate of gastric emptying of liquids, with few exceptions (Schulze-Delrieu and Brown 1985). Unfortunately, the stonts used in these studies projected into the antrum and duodenum, and may have affected motilily in these regions as well as the pylorus.

The inconsistency of these results may be related to differences in methodology, such as the use of different test meals, variations in the length and orientation of the pyloroplasty, and unequal post-prandial recording times. Furthermore, no attempt was made to control for other motor mechanisms (see Section 6.1). The potential for compensation by these other mechanisms is high- lighted by the results of combined pyloroplasty and truncal vagotomy, which almost universally has resulted in acceleration of liquid gastric emptying (MacGregor et al 1977; Hinder and Bremner 1978; Yamagishi and Debas 1978; Gulsrud et al 1980; White et al 1984), presumably because compensatory changes in antral or fundal motility were prevented by vagal denervation. ..+ rrt

There is limited information on the mechanical importance of localized pylor¡c contraction in humans. Concurrent measurements of gastric emptying and pyloric motility indicate that the slow gastric emptying of nutrient-rich meals of flavoured milk and 25I" dexlrose is associaled with increased numbers of isolated pyloric pressure wavos (lPPWs) (Houghton et al 1988aab) (see Sections 4.2.2 and 4.3.2). The demonstration in humans by Tougas et al (1987) that IPPWs and pyloric tone induced by inlraduodenal lipid are associated with the cessation of transpyloric flow of liquid barium, perhaps provides the strongest evidence to date that localized pyloric contraction retards the gastric empty¡ng of nutrient liquids.

Before doubts about the importance of the pylorus in the regulation of liquid gastric emplying can be finally laid to rest, additional, carefully designed studies in animal models and humans are required. The debate over the effect of pyloroplasty may be settled by prolonged studies of the effect of short, axial pyloroplasties, with complete division of the distal pyloric muscle loop, on the emptying of nutrient and non-nutrient test meals. Other important motor mechanisms should be controlled for where possible. These considerations apply equally to studies examining the

I effect of pyloric exoision. ln humans, substantial insight into the mechanical importance of localized pyloric contractions would be gained from further concurrent measurements of intraluminal pressure and gastric emptying or lranspyloric flow. ! 62

6.3.2 The Pylorus and Gastrlc Empty¡ng of Dlgestlble Sotids The role of the pylorus in conlrolling solid gastric emptying is perhaps evên more difficult to establish, because of the additional complexity of this process. Following ingestion, solids are inítially retained in the proximal stomach. Before gaslric emptying begins, there is an initial lag phase (Figure 3.2), during which solid food is redistributed from the proximal gastric reservoir to the distal stomach (Collins et al 1988). There is some evidence that larger particles may be preferentially retained in the proximal stomach as redistribution occurs, so that the proximal stomach may play a significant role in solid-liquid discrimination and sieving of solid food (Meyer 1987). Whether the pylorus participates in lhese early processes is unknown, but, conceivably, resistance to transpyloric flow provided by the pylorus could be important (below). ln addition to redistribution, the lag phase is characterized by further processing of the solid in the distal stomach by propulsion, grinding and retropulsion by antral contractions, until solid particles have been sssentially liquified (. 1mm in diameter) (Thomas 1957; Meyer et al 1979; Meyer 1987). The subsequent emptying of solids is thought to depend on continued effective grinding (Cole 1928; Brody et al 1940; Kelly 1980; Camilleri et al 1985). Once this has occurred, the mechanisms controlling empty¡ng of the 'liquified' solid are probably similar to those controlling emptying of nutrient liquids of similar caloric or nutrient densily (Hinder and Kelly 1977; Meyer 1987), which empty at a similar rate when ingested simultaneously (Seigel et al 1988).

4 Cannon (1898) ''i; observed fluoroscopically in cats that solid particles were stopped at the pylorus j by contraction of the pylor¡c sphincler, and then retropulsed. Similar observations have been rePorted by others (Hoffmeister and Schutz 1886, Klein 1926, cited by Thomas 1957). Retention of solids by pyloric closure followed by retropulsion is a possible mechanism which would permit grinding of solids by antral contractions without premature emptying of large particles. lsolated pyloric pressure waves (lPPWs) are observed during solid emptying (Houghton et al 1988aab; Heddle et al 1989) (Figure 6.1), but a relationship between the number of phasic pyloric contraction and the rate of gastric emptying has not yet been established, and there have been

few satisfactory measurements of tonic pyloric contraction during the emptying of a solid meal. Houghton et al (1988b) observed that increased numbers of lPPWs, after ingestion of a mixed meal containing a liquid dextrose component, were associated with a prolonged lag phase before the start of solid emptying (Figure 6.1). lnlerestingly, however, liquid continued to empty during the solid lag phase, presumably in the intervals between localized pyloric contractions. This may reflect sieving of the most liquid contents through a relatively narrow pylorus (Ehrlein 1988).

The motor mechanisms underlying sieving, or solid-liquid discrimination, are poorly understood,

I but the process could be explained by a simple fluid mechanics model of fluid and particle flow, in

which particle size is the major determinant of how the particle moves relative to the fluid in which it is suspended (Amidon 1985). Thus, during antral contractions, as fluid is forced down I 63

the cone-shaped antrum, very small part¡cles are swept down the centre of the lumen and ejected through the pylorus, wherêas larger particles remain near lhe antral wall, only lo be pushed retrogradely when the pylorus closes for further grinding. The results of several experiments suggest that the pylorus may actively participate in this process. Fluoroscopic observations by Dozois et al (1971) and Wilbur and Kelly (1975) showed that indigestible plastic spheres were retained in the stomach during the rapid emptying of liquid. This sieving was associated with simultaneous contraction of the pylorus and terminal antrum which resulted in retropulsion of the antral contents.

Surgical interference with the pylorus has been shown to impair normal solid-liquid discrimination. Meyer et al (1979) performed meticulously designed experiments in several groups of mongrel dogs, and compared the effects of truncal vagolomy, pyloroplasty, pyloroplasty and vagotomy, and antrectomy alone. The longitudinal pyloroplasty extended into the antrum and duodenum for 1.5 to 2.0 cm either side of thê pyloric ring. Dogs were fed a solid meal consisting of 1cm cubes of liver and beefsteak, together with a variable quantity of water. Analysis of the gastric effluent collected from a duodenal cannula showed that the majority (95%) of particles emptied in control animals was less than or equal to 0.25 mm in diameter, with only 1.3% being more than 2 mm in diameter. Pyloroplasty led lo some impairment of solid- liquid discrimination (5.3% >2 mm diameter), but far greater effects were seen after :'i rLt .], pyloroplasty and vagotomy, and Bilroth ll antrectomy (25.8% and 329." >2 mm diameter respect¡vely). Hinder and San-Garde (1983) showed that, in contrast to antrectomy alone, pylorectomy plus antrectomy resulted in 38% of solid emplying occurring as particles >1mm in diameter, compared with only 2.5"/" in control animals. Dozois et al (1971) found that large indigestible spheres 1cm in diameter emptied after pylorectomy. Such particles normally emply during antral phase lll of the interdigestive migrating motor cycle (see below).

lnterference with the pylorus in animal models has produced inconsistent effects on the rate of

gastric emptying of nutrient solids. Pyloroplasty alone, or combined with antrectomy, has not been shown to alter the rate of solid gastric emptying (Yamagishi and Debas 1978; Meyer et al 1979). The results of combined pyloroplasty and vagotomy on rates of solid emptying are inconsistent, with one study showing an increase (MacGregor et al 19771, and another showing no change (Meyer et al 1979). Pylorectomy, however, may result in an increased rate of empty¡ng of solids in the pig (Treacy et al 1988a). Pylorectomy also leads to increased rates of solid emptying when combined with antrectomy (Hinder and San Garde 1983), whereas t I antrectomy alone has little effect (Meyer et al 1979; Hinder and San Garde 19gg).

; The pylorus is likely to be as important in the emptying of 'liquified' solids as it is with nutrient liquids. The pylorus may also have a significant role in the initial grinding and sieving of solid

r 64

¡üt A r80 IÉ Eco

N- Sellnê Solld Eo € 4 tæ

0 n 0 tãl ltt ¡80 2t0

''i il iu$ill'ryllrftl iljilfilfiflur r#ïil fifliililhlr Hilh ¡J i tlilil

0 r¡ 60 s lz0 1t0 lE0 ?10 Ilnc lrÍnl

cl0 a Ê !o ?Í De¡los: Solld 5 Bao I

x 6 m læ ¡tt 180 2t0 ao

6 tt5¡ ji#fi| D il rï' r;r;;' illt ilil, ;ii tTrl;ill;,ilt, ï1;tir,ïtili; 0 qt x ó0 I 2t! tr0 It0 ?10 20 I¡m! lnlnl

Figure 6.1 Gastric emptying profiles in normal subjects for solid and liquid components of mixed meals consisting of (A) 200m1 normal saline and a 1009 chicken liver/ground beef burger, and (B) 200 ml 25 "/" dexlrose with an identical burger. The lower panels represent different types of contraclile activity in the pylorus (S), stomach (G) and duodenum (D). Pressures are measured wilh a sleeve sensor (pylorus) and side-holes (3 antral, 4 duodenal) spaced at 3 cm intervals above and below the sleeve. The solid lines represent coordinated pressure waves, and the dashed lines represent random or isolated pressure waves. The dofs represent lPPWs. The onset of solid emptying was associated with an increase in coordinated events involving the pylorus together with the proximal antrum. Note in particular lhat a prolonged solid lag phase occurred I after ingestion of the dextrose liquid meal, and persisted until most of the l¡quid had emptied. This period was associated with frequent isolaled pyloric pressure waves, and few propagated waves. (From Houghton et al 1988b, with permission.)

* 65

food. Many questions, however, remain unanswered. Further studies in carefully designed animal models are required, with appropriate measurement and control of other motor mechanisms where technically feasible. ln humans, accurale correlations of the numbers of phasic pyloric contractions and the degree of pyloric tone with the rate of gastric empty¡ng should be performed, particularly under conditions where the rate of emptying is changing.

6.3.3 The Gastrlc Emptylng ol Non-Digestible Solids Durlng Fastlng No studies have examined the effect of pyloric motility on the emptying of non-digestible solids, which normally empty during phases ll and lll of the fasting migrating motor cycle (Carlson et al 1966; Code and Schlegel 1974; Hinder and Kelly 1977). While the pylorus may be critical for the retention of such particles during fed motility (Dozois et al 1971; see above), it probably only has a minor independent role in controlling their final emptying, since the incidence of localized pyloric contraction during fasting is low (Chapter 4). The pylorus is likely, however, to participate in propulsion as part of the antral pump. This deserves specific investigation.

6.3.4 The Pylorus and Duodenogastric Reflux The occurrence of duodenogastric reflux is thought lo be primarily due to lhe Pattêrn of duodenal motility, such as synchronous or retrograde contractions (Johnson and Eyre-Brook 1984), but it may also be related to the liming of duodenal with pyloric contractions (Johnson 1975; Muller- Lissner and Schindelbeck 1984; Chaussade et al 19S9). The suggestion that the primary role of the pylorus was to prevent duodenogastric reflux was based on rePorts that pyloric contractions may coincide with those of the duodenal cap (Meschan and Quigley 1938; Thomas 1957; Fisher and Cohen 1973). Studies using intraluminal electrodes, however, suggest that the pylorus rarely contracts with or just after the duodenum (Johnson et al 1983). All the above studies are of questionable validity (Chapter 2). Fluoroscopy of canine gastric motility during induced vomiting suggests that reflux does not occur when the pylorus is closed (Code et al 1984), but the significance of this is also hard to interpret because of the use of intravenous opiates and electrical pacing. Thus, although some evidence favours a role for the pylorus in preventing duodenogastric reflux, additional, technically adequate studies need to be performed.

6.4 CONCLUSION Non-localized pyloric contractions may participate in coordinated antroduodenal pumping. lsolated pyloric pressure waves, however, are likely to be non-propulsive and to provide a resistance to transpyloric flow. Localized pyloric contraction may play a significant role in slowing the rate of emptying of nutrient liquid and liquified solid meals, and in the grinding and sieving of large solid particles. Further studies are required to investigate these functions in more detail, including under conditions of altered gastric emptying other than during upper small intestinal receplor stimulation. The role of the pylorus in preventing duodenogastric reflux remains unclear. 66

CHAPTER 7

Methods Used to Study Pyloric Mottlity and Gastric Emptying in This Thesis

7.1 INTRODUCTION The studies reported in Chapters I to 10 were designed to investigate aspects of pyloric motor function and its measurement in humans. All studies were performed in healthy volunteers. The primary methodological requirements for these studies were that pyloric activity should be reliably recorded, and that the estimation of gastric omptying should be performed both accurately and without interference to concurrent molility recordings. Furthermore, neither technique should itself induce significant alterations of gastrointestinal function. lntraluminal perfused sleeve and side-hole manometry was used in all studies to record pyloric and antroduodenal molility respectively. Gastric emptying was measured with a radionuclide technique. These methods are established and previously validated techniques, and were chosen as the most appropriate based on the factors considered in Chapters 2 and 3. ln this chapter, the methods common to all studies will be described, providing information required for the proper interpretation of the studies and their results. Variations in these methods were required because of the specific demands of some studies, and these new developments will be specifically discussed in the appropriate chapters (see especially Chapter 8).

7.2 SELECTION OF SUBJECTS Studies were performed in healthy, adult volunteers, who were recruited from the campus of a local university and by advertising within the Royal Adelaide Hospital. Each volunteer underwent a brief medical assessment to exclude those with a present or past history of significant dyspepsia, peptic ulcer or other gastrointestinal disease. Those with a history of upper 67 gastro¡ntestinal surgery were also excluded, as were those taking any drugs which could inlerfere with gastrointestinal motility. Subjects with structural nose or throat abnormalities were not considered for intubation. ln those studies involving fluoroscopy or radionuclides, females who were or could have been pregnant were also excluded.

Volunteers were fully informed of all aspects of experimental protocols and were provided with wrítten information detailing the procedures to be used, as well as any potential risks from pharmaceuticals or radiation where relevant. All volunteers signed a consênt form prior to starting the experiments. Study protocols were separately submitted to, and approved by, the Human Ethics Committee of the Royal Adelaide Hospital during 1987.

7.3 MANOMETRIC TECHNIOUE 7 .3.1 lntubatlon Technique Subjects fasted overnight and were requested not to consume alcohol the night before, nor to smoke on the day of the study. All studies commenced at approximately 0900 in the morning. The manometric catheter was introduced transnasally, after one nostril had first been anaesthetized with lhe local anaesthetic lignocaine. Passage of the catheter past the upper oesophageal sphincter was aided if necessary by small sips of water through a straw. After intubation, the subjects lay on the right side to facilitate passage of the terminal end of the catheter across the gastroduodenal junction. Entry into the duodenum normally occurred during propagated phase ll activily of the interdigestive migrating motor cycle, and was indicated by observing typical fasting duodenal pressure waves in the side-holes, as well as by a change in the lransmucosal potential difference measured from the side-hole at the distal end of the sleeve, from an antral to a duodenal value (see below). The catheter was then positioned with the sleeve sensor across the pylorus, astride the gastroduodenal transmucosal potential difference gradient. Crileria for correct placement of the sleeve are described below.

7.3.2 Manometrlc Methods 7.3.2.1 Design of the Manometric Assembly The manometric assemblies in all studies were made within the Gastroenterology Unit of the Royal Adelaide Hospital. Designs were tailored according lo the requirements of each study. Manometric assemblies were constructed from silicon rubber, and incorporated a sleeve sensor together with an array of perfused side-holes in the antrum and duodenum (Chapter 2). With one exception (see Section 8.3), the sleeve sensor in all studies was 45 mm in length and 5 mm in maximal external diameter. This sleeve length was found in previous experiments (Heddle et al 1988c) to adequately tolerate axial movements relative to the pylorus wilhout protruding excessively into the duodenal cap (Chapter 2). lmmediately beyond the sleeve, the calheter was very flexible, in order lo accommodate the curvature of the proximal duodenum without causing 68 leverage on the sleeve segment. The sleeve was posilioned across the gastroduodenal transmucosal potential difference (TMPD) gradient (see below) and was used to record changes in intraluminal pressure at the pylorus. ln these assemblies,5 side-holes were distributed at 15mm intervals on the reverss asp€ct of the sleeve, from a point level with the proximal end to a point level with the distal end of the sleeve. These side-holes were used to determine the propagation characteristics of pressure waves, and to help differentiate localized pyloric contraction from less localized waves which also involved the terminal antrum or proximal duodenum (Chapters 2, 4 and Section 8.2). The side-holes at either extrsmity of the sleeve recorded both intraluminal pressures and transmucosal potential difference, and are referred to as the antral and duodenal TMPD side-holes. Antral and duodenal pressures were also recorded from two side-holes located 2 and 4 cm proximal to the sleeve in the anlrum, and lwo side-holes 3 and 9 cm distal to the sleeve in the duodenum (Figure 7.1). A tungsten weight was incorporated at the distal end of the catheter to help position the catheter in the lerminal anlrum prior to passage through the pylorus and to help maintain the end of the catheter around the curue of the duodenum during prolonged manometric studies. Manometric assemblies used in the experiments reported in Chapters 8.2, 9.2 and 10 were identical. A schematic diagram of their design is shown in Figure 7.1. The assemblies used in the remaining two experiments are described in Chapters 8.3 and 9.3.

7.3.2.2 Recording of lntraluminal Pressures Manometric channels were perfused at a constant rate of 0.3 ml/min, wilh a low compliance pneumohydraulic pump driven by a reservoir pressure of 50 KPa. The two side-holes at either end of the sleeve, which were used to measure lransmucosal potential difference, were perfused with degassed normal saline from separale reservoirs. All other channels were perfused with degassed water from a third reservoir. Pressures were recorded with prêssure transducers (Deseret Medical lnc. cat no.38-8000-1, Sandy, Utah, USA) interfaced to a 12 channel chart recorder (Grass model 7D, Grass lnc. Quincy, Mass, USA), which was run at a paper speed of 100 mm/min (Figure 7.2). Bench testing showed that occlusion of the sleeve sensor caused a pressure rise rate of 150 mmHg/sec at the proximal sleeve, 12 mmHg/sec at mid-sleeve and 6 mmHg/sec at the distal sleeve. This substantial difference in pressure rise rates is explained by the compliance of the sleeve membrane (Section 2.3.1.3). The rate of change of basal pyloric pressure (in the order of minutes rather than seconds), and the duration (approximately 6 to 10 seconds) as well as the amplitude (frequently less than 40 mmHg; see, for example, Figures 4.3 and 9.2) of isolated pyloric pressure waves, are such that the slow rise rates in the distal half of the sleeve are not critical and should not result in any significant under-recording of pyloric motility. Nevertheless, during each sludy, the assembly was posilioned whenever possible so that the zone of pyloric contraction was at, or orad to, the mid-sleeve. The pressure rise rate of each side-hole was 160 mmHg/sec, less than the rise rates often produced by standard oesophageal manometric side-holes. This difference is explained by the longer catheter and lower perfusion rates and pressures used in the present experiments. 69

SLE

TMPD

SIDE

WEIGHT

Flgure 7.1 Schematic representation of the manometric assembly positioned across the pylorus. Pyloric pressures are recorded by the sleeve sensor, which is maintained across the gastroduodenal junction with the aid of feedback from contlnuous measurements of transmucosal potential difference (TMPD) from the side-holes at eilher end of the sleeve. Antral and duodenal pressures are recorded from additional side-holes sited at intervals along the catheter and on the reverse aspect of the sleeve. 70

7 .3.3 Measurement of Transmucosal Potential Dlfference 7.3.3.1 Background The electrical potential difference which exists between the serosal and mucosal surfaces of the gastrointeslinal tract is thought to be a function of ionic flux across the mucosa (Adams and Davies 1963). Potential difference measurements have been used in the diagnosis of oesophagitis (Khamis et al 1978; Orlando et al 1982) and as an index of the function and integrity of the human gastric epithelium (Geall et al 1970; Cooke 1976). Differences in mucosal properties result in a substantial difference between the transmucosal polential difference (TMPD) in the distal antrum and the proximal duodenum, so thal a sharp gradient exists at the pylorus (Andersson and Grossman 1965; Chapter 1). This observation has been recently exploited in humans to help localize manomelric assemblies at the gastroduodenal junction (Fisher and Cohen 1973; White et al 1981 & 1983; Tougas et al 1987; Heddle et al 1988a,bac; Houghton et al aab; see also Chapter 2).

The transmucosal potenlial difference measured by intraluminal side-holes can be influenced by mucosal abnormalities, as well as by unusually high or low gastric acidity, which may cause substantial electrical potentials at the junction between the electrode and gastric juice (Read and Fordtran 1979). ln practice, the effect of excessive junction potentials is minimized by using a flowing column as the intraluminal electrode and is largely balanced by a subcutaneous reference eleclrode (Read and Fordtran 1979). Normal saline is the preferred conducting solution for the flowing eleclrode, mainly because of its safely. These principles were employed in the method developed by Heddle, Dent and co-workers (Heddle 1989), which enables predictable and stable TMPD moasurements in normal humans. This method has been used successfully in recent manomelric studies in humans (Heddle et al 1988a,bac; Houghton et al 1988aab)

7.3.3.2 Technique The technique used to measure transmucosal potential difference (TMPD) was based on that of Heddle (1989). The side-holes at either end of the sleeve sensor were perfused by flowing normal saline columns, which made electrical contact with bridges filled with 1M KCI in 3/" agar via rigid Y connectors at the proximal end of the manometric catheter. The other ends of the bridges were connecled to calomel half cells (lonode, Brisbane, Australia). The common reference electrode was a sterile plastic cannula inserted subcutaneously on the back of the volunteer's right forearm. This was connected via a KCI-agar bridge to a third calomel half cell. Calomel half cells were balanced at the start of each sludy so that the voltage asymmetry, or back-to-back potential, was less than 2 mV.

TMPD measurements were recorded continuously on the chart recorder throughout each experiment, and the manometric assembly was withdrawn or advanced as necessary to keep the sleeve sensor astride the gastroduodenal TMPD gradient. Criteria for correct positioning of 71

Figure 7.2 Photograph of an intubated subject seated in front of the gamma camera (a) for a combined motility/gastric emplying study. Scintigraphic images can be observed on a cathode ray oscilloscope and are processed by a computer. The manometric catheter (b) is perfused by an adapted pneumohydraulic pump, housed together wlth the calomel cells and agar bridges within a Faraday cage (c), to minimize external electrical interference with TMPD measurements. TMPD is indicated on meters (d) and recorded continuously with the pressure tracings on lhe Grass polygraph (e). 72 lhe sleeve were similar to those employed in recent studies by Heddle et al (1988a,b&c) and Houghton et al (1988aab) and were that: 1) antral TMPD should be more negative than -20 mV; 2) duodenal TMPD should be more positive than -15 mV; 3) the difference between antral and duodenal TMPD should be at least 15 mV.

7.4 MEASUREMENT OF GASTRIC EMPTYING Gastric emptying was monitored with a radionuclide technique (Chapler 3). Subjects ingested a solid meal, which consisted of 100 grams of cooked ground beef containing chicken liver labelled in vivo on the morning of the experiment with 37-55 MBq (1-1.5 mCi) ee'Tc-sulphur colloid (Meyer 1976). The test meal contained 259 protein,2lg fat, and had a calorie contenl of 27o kcal. The meal was consumed wilhin 5 minutes with 30 ml of water.

Subjecls were seated comfortably on a stool wilh their arms resting on a table (Figure 7.2). Gaslric empty¡ng was monilored with a single, posteriorly positioned gamma camera (Nuclear Chicago Pho-Gamma 111 HP, Digital Equipment Corporation), with a high energy 400 keV, 4000 parallel hole collimator. Scintigraphic image data was interfaced to a computer (PDP 11/55). Data were collected at a frame rate of between 1 and 3/min. The gastric image was displayed on the computer scrsen, and lines were constructed around gastric, proximal gastric and distal gastric regions of interest (ROl) with the aid of a cursor. The proximal stomach region was defined as lhe "reservoir" area seen in all subjects during the first few minutes after ingestion of the solid meal, and a horizontal line was drawn immediately below this region to separate il from the distal gastric ROI (Collins et al 1984). Activity time curves were derived for each of the three ROls (proximal, distal and total stomach). For each image, counts were corrected for subject movement and radionuclide decay (Chapter 3). The error due 1o radionuclide gamma ray attenuation was minimized by the use of correction factors derived from a lateral image of the stomach (Collins et al 1983; Section 3.3.5.4). At approximately 30 minute intervals, data acquisition was interrupted for 1-2 minutes while computer disks were changed and, in order to provent fatigue, the subject was allowed to stand or sit away from the camera during lhis time.

7.5 DATA ANALYSIS 7.5.1 Analysis of lntralumlnal Recordings Records were only analysed when the lransmucosal potential difference measurements indicated correct positioning of the sleeve across the pylorus, according to the criteria outlined under Section 7.3.9.2.

With the exception of the study described in Chapter 8.3, pressure waves recorded in all side- holes, as well as in the sleeve, were scored only if their amplitude was greater than or equal to 73

10 mmHg. This cut off, used in previous sludies of pyloric motility (Heddle et al 1988a,bac; Houghton et al 1988aab), was designed to prevent inclusion in the analysis of small fluctuations in pressure due to respiratory oscillations, straining or artefacts. The criteria used for scoring in Chapter 8.3 are described in detail in Section 8.3.3.1. Pressure waves recorded in lwo or more adjacent side-holes were judged to be propagated antegradely if the onset of the major initial upstroke occurred >0.5 sec and <5 sec A[leI lhat of a similar pressure wave recorded in the immediately orad side-hole. lf the major initial upstroke occurred >0.5 sec but <5 sec þgþle that in the immediately orad side-hole, retrograde propagation was judged to have occurred (Houghton et al 1988aab). Pressure waves demonstrating this close temporal association were considered to be coordinated (Chapter 4). ln some studies (Chapters 9 and 10), the number of contractions recorded in all antral and duodenal side-holes were summed to provide a simple motilily index for each 5 minute interval.

Waves greater than or equal to 1OmmHg in amplitude which were recorded by the sleeve spanning the pylorus were classified as isolated pyloric pressure waves (lPPWs) according to the criteria of Heddle, Dent and co-workers (Heddle et al 1988a,bac; Houghton 1988aab). These criteria were that the sleeve-recorded prsssure waves should be: 1) narrow, ie. recorded by no more than 1 side-hole along the length of the sleeve; 2) recorded in the absence of an associated wave of any amplitude in the antral or

duodenal TMPD side-holes. The duration of the IPPW response, where relevant, was defined as the period for which the mean number of IPPWs/min in the test period remained greater than lhe range observed in the immediately preceeding control period.

Basal pyloric pressure was determined for each minute of the study as follows: the mean end- expiratory baseline pressure recorded by the sleeve sensor was determined visually, by drawing a line below the respiratory prêssure oscillations for each minute of the sludy, after first editing out phasic pressure evenls and artefacts; the mean end-expiratory pressure recorded simultaneously by the most distal antral side-hole was determined in the same way; the difference between the sleeve-recorded and the distal antral baseline pressure was calculated after correcting for differences in horizontal perfusion pressures determined for each channel by bench-testing before the start of each study. The difference between baseline sleeve and distal antral pressure was called the basal pytorlc pressure or pylorlc tone (Chapter 4), and was expressed as the mean basal pyloric pressure in mmHg for each 5 minute study period. The duration of a pyloric tonic response was defined as the number of consecutive minutes that basal pyloric pressure was >2 mmHg. 74

Lag Phase

,-aaa\ a Retentlon at 60 mln

Post- la g Llnear Emptying Rate

I o O/" RETENTION a b G

c d

0 20 40 f(x) 120 POST.CIBAL MINUTES

Flgure 7.3 Schematic representation of the analysis of a radionuclide gastric emptying curve of the emptying of a 1009 chicken liver/ground beef burger in a normal volunteer. ln this example, the lag phase is approximately 20 min, the post-lag linear emptying rate (calculated from [a-b]/[d-c]) is 1%/min, and lhe retention at 60 min is 660/". 75

7.5.2 Analysls of Gastrlc Emptylng Measurements

Gastric emptying cules were analysed by a method described in detail by Collins et al (1983 & 1984). Primary analysis of gastric emptying was based on the isotopic counls delected within a region of interest drawn around the total stomach (Section 3.3.5.3). A number of parameters were used to define the characteristics of the gastric emptying curve (Figure 7.3).

The lag phase was defined as the time taken before any of the meal emptied from the stomach. The@ing(expressedaS7"emPtied/min)wasdeterminedvisuallyby drawing a line of best fit through the points on the curve, subsequent to the end of the lag phase. The amount of the meal remaining in the stomach (7" retention) at predetermined intervals, for example 60 minutes after meal completion, was determined in some of the studies. ln lhe studies described in Section 9.3 and Chapter 10, the regional distribution of the solid meal within the stomach was also determined by constructing proximal and distal gastric regions of interest (Collins et al 1988). The proximal region of interest was defined as the 'reservoir" area observed for at least the first few minutes (Section 7.4). The distal gastric region of interest was then defined as the remaining, distal porlion of the total gastric region of interest. The line of proximal/distal subdivision usually closely approximated the position of a midgastric band (Moore et al 1986).

7 .5.3 Statistlcal Tests Data were analysed using the Wilcoxon matched pairs signed-rank test. Unless olherwise staled, two-ta¡led tests of significance were performed. Correlations were performed with Spearman's correlation coefficient. A p value of <0.05 was considered significant in all analyses. All values are expressed as medians and interquartile ranges unless otherwise stated. 76

CHAPTER 8

Technical Aspects Involved in Interpreting Combined Sleeve/Stde-Hole Pylorlc Manometry in Humans

8.1 INTRODUCTION ln this chapter the results of two sets of experiments, which were designed to further investigate technical issues relevant to sleeve/side-hole manometry in humans, are described. An improved understanding of these issues was oxpected to aid the interpretation of manometric recordings. Antropyloroduodenal motility was initially studied under different conditions of gaslrointeslinal intubation, in an attempt to resolve an existing conflict in the literature about the influence of intubation ilself on gastrointestinal function. ln the second set of experiments, queslions concerning topographical patterns of pyloric motility, especially in relationship to antral contractions, were explored with the aid of a novel sensor which detected antral contractions with greater sensitivity than antral manometry.

8.2 A STUDY OF THE COMPARATIVE EFFECTS OF DUODENAL AND ILEAL INTUBATION ON PYLORIC, ANTRAL AND DUODENAL MOTILITY, AND THE GASTRIC EMPTYING OF A SOLID MEAL 8.2.1 Background The influence of gastrointestinal intubation on gastric and duodenal motility is of fundamental importance to the interpretation of many tests of gastrointestinal function, including pyloric manometry. Available informalion about this issue is limited and conflicting. Muller-Lissner et al (1982) reported that the prêsence of small gastric and lranspyloric tubes did not affect gastric emptying of a fat-containing liquid meal. Similarly, Longstreth et al (1975) found that a transpyloric tubs did not alter gastric emptying of a solid meal. ln apparent contrast to these observations, the study of Read et al (1983) indicated that the presence of an intestinal tube 77 ending in the ileum resulted in significant slowing of gastric emplying and, somewhat paradoxically, in more rapid small intestinal transit. No study, however, has compared the effect of gastroinlestinal intubalion with calheters of different lengths on gastric emptying of identical meals. Moreover, the gastric and pyloric motor rêsponses lo differing techniques of gastrointestinal intubation have not been measured. ln this study, simultaneous measurements of pyloric, antral and duodenal motility and gastric emptying were employed to evaluate the comparative influence of gastric and small intestinal intubation on pyloric motility and gastric emptying of a standardized solid meal.

8.2.2 Materlals and Methods 8.2.2.1 Subjects Studies were performed on 26 healthy volunleers (18 men, I women) aged between 19 and 26 years. ln 10 subjects (6 men,4 women), a catheter was passed as far as the fourth part of the duodenum (Group D); in I subjects (5 men, 3 women) the terminal ileum was intubated (Group l); the remaining 8 subjects (7 men, 1 woman) were not intubatsd (Group C).

8.2.2.2 Experimental Protocol ln these studies, lwo manometric assemblies were used, which were virtually identical apart from their length. Each assembly included a sleeve sensor of similar dimensions (5 mm wide, 45 mm long) for measurement of pyloric pressures, and a series of perfused side-holes to record pressures in the antrum and duodenum (Chapter 2). One catheter ended in the duodenum near lhe ligament of the Treitz (Group D), whilst the other catheter extended beyond the assembly for 150 cm and incorporated an infusion port and a silicon rubber balloon at its most distal end

(Group l).

Subjects were intubated in a standard manner at 0900 hours (Chapter 7), and the end of the calheter was positioned in the duodenum. When this had been achieved, passage of the tip of the longer manometric assembly through the small bowel was stimulated by inflation of the balloon with 10 ml of air. This catheter was then allowed to pass down the small intestine until its distal end was 210 cm from the lip of the nose, when the balloon was deflated. The siting of the end of ths catheter in the terminal ileum was confirmed by fluoroscopy. The position of both manometric assemblies was then finally adjusted until the sleeve straddled the pylorus, with the aid of feedback from measurements of transmucosal potential difference (TMPD) from the side- holes at either end of the sleeve (Chapters 2 and 71.

When the manometric assemblies were correctly positioned, the subjects sat with their back to a gamma camera, and normal saline (NS) was infused at 1 ml/min into either the duodenum (Group D), or the ileum (Group l). After 10-15 minutes, each subjecl was given the standardized solid 78 test meal described in Chapter 7. The time between the commencement of ¡ntubation and the ingestion of the test meal (intubation time) was recorded. Gaslric emptying and pressures in the antrum, pylorus and duodenum were monitored continuously until the end of each study. ln order lhat the gastric emptying measurements would commence at approximately the same time of day in each of the three study groups, the I subjects who were not intubated were given the lest meal somewhat later than the start of the other experiments, at 1000 hours.

8.2.2.3 Data Analysis Records were analysed for four 15 minute periods commencing immediately after the lag phase of the meal. ln each study, pressure waves with an amplilude >1 0 mmHg were counted for the antral TMPD side-hole, the antral side-hole 2 cm orad to the sleeve sensor, the sleeve sensor straddling the pylorus and for thê duodenal side-hole 4 cm aborad to the sleeve. During each analysis period, the mean rate of pressure waves/min was calculated for each recording point. Pressure waves recorded by the sleeve sensor were divided into isolated pyloric pressure waves (lPPWs) and non-localized pressure waves. Basal pyloric pressure was referenced to basal antral pressure recorded by the side-hole 2 cm proximal to the sleeve sensor.

Data in this study were assessed using the Mann-Whitney-U-test ancj iinear regression analysis, and are expressed as medían values with ranges unless otherwise stated. A p value of <0.05 was considered significant in all analyses.

8.2.3 Resu lts Nasogastric intubation and ingestion of the test meal were well tolerated and no subject reported nausea. The sleeve sensor was located correctly across the pylorus for more than 90% of total recording lime. As would be expected, the time between the commencement of intubation and ingestion of the tost meal (intubation time) was substantially longer in the ileal compared 1o the duodenal inlubation group (365 min [300-525] vs 87 min [55-108], p<0.001).

8.2.3.1 Gastric Emplying ln all subjects the rate of gastric emptying of the test meal was close to linear, after an initial lag phase. There was no significant difference in measures of gastric emptying between the control and duodenal intubation groups. ln the ileal intubation group, the time before any of the solid meal emptied from the stomach (lag phase) was not significantly different from the two other groups, but the linear emplying rate was slower (p<0.01) (Figure 8.1), and the amount of the meal remaining at 60 minutês was greater (p<0.05), than in the other two groups (Table 8.1). 79

2.0

1.5 a

a LINEAR a EMPTYING a 1.0 a RATE a a a

o/"lmln a a a I a a a a 0.5 a - l-t *

0.0 Control Duodonal llea I Group lntu batlon I ntu batlo n

Flgure 8.1 eetTc-labelled lndlvidual values for the post-lag linear emptying rats of a solid meal in control, duodenal intubation and ileal intubation groups. The median values are indicâted by the horizontal bars. *p<0.01 compared to control and duodenal lntubatlon groups. 80

Parameter Control duodenal tube ileal tube

number of subjects I 10 I lag phase (min) % (1747') 39 (1047) 38 (12€1)

post lag linear 0.82 (0.63-1.40) 0.76 (0.43-1.81) 0.4 (0.36-0.90)" emptying rate (%/min)

retention at 60 min (%) 76 (s1-e4) 79 (s7-e1) gg (73-1OO)"

Table 8.1 Measures of solid gastric emptying in control subjects and after duodenal and ileal intubation. Data are median values with ranges in parentheses. " p<0.05 compared to control and duodenal tube groups.

Parameter (waves/min) duodenal lube ileal lube p value

antrum 0.2 (o-1.8) 0.02 (o-2.0) n.s. terminal antrum 1.7 (0.s2.5) 0.8 (o-2.2) o.o2 IPPWs 0.5 (0.+2.0) 0.8 (0.2-1.r) n.s.

duodenum 1.7 (0.7-2.3) 1.7 (0.4-2.4) n.s.

Table 8.2 Antrel, pyloric and duodenal motility after duodenal or ileal intubation. Pressure waves were recorded from the side-hole 2 cm orad to the sleeve (antrum), the antral TMPD side-hole (terminal antrum), the sleeve sensor (lPPWs) and the side-hole 4 cm distal to the sleeve (duodenum). Data are median values with ranges in parentheses.

8.2.3.2 Molilitv These results are summarized in Table 8.2. There were significantly fewer (p<0.01) distal antral pressure waves recorded by the antral TMPD side-hole in the ileal compared to the duodenal intubation group, and a trend towards fewer antral pressure waves at the more proximal antral recording site 2 cm above the sleeve (Figure 8.2). The higher median rate of IPPWs in the ileal intubation group did not reach significance. Basal pyloric prossure showed no difference between the two groups. There was no significant difference in the number of duodenal pressure waves between the two groups. 81

3.0

2-5 a

I a 2.O a a a ANTRUM wavcs/mln 1.5 aa

a 1.0 a a a t a 0.5

& T aa 0.0 a

Duodcnal llca I lntubatlon lntubatlon

Flgure 8.2 lndivldual valuss for the number of antral pressure waves recorded by the entral TMPD side-hole in duodenal intubation and ileal intubation groups. The median values are shown (horlzontal bars). rp

8.2.3.3 Relationships Between Gastric Emptying. Motilily and lntubation Time There was a significant inverse correlation between the number of waves recorded by the antral TMPD side-hole and the lime between starting intubation and ingestion of the meal (r= -0.69, p<0.01). The inverse relationship between the intubation time and the linear gastric emptying rate was not statistically significant (r= -O.42, p<0.1). The relationships between the number of pressure waves recorded by the antral TMPD side-hole and both the amount of isotope remaining in the stomach at 60 minutes (r=0.30) and the linear emptying rate (r=0.14) were not slatistically significant.

8.2.4 Summary of Results These experimenls compared the fed patterns of pyloric, antral and duodenal motility following transpyloric intubation of the duodenum and terminal ileum, as well as lhe rate of gastric emptying in intubated and non-intubated subjects. There was no significant difference in pyloric motor patterns in the duodenal, compared with the ileal intubation group. ln contrast, gastric emptying in the ileal, but not the duodenal intubation group, was substantially slower than both the duodenal intubation group and the non-intubated group. There were significantly less antral pressure waves in the ileal intubation group compared to the duodenal intubation group.

8.3 AN EVALUATION OF THE NORMAL PATTERNS OF HUMAN PYLORIC AND ANTRAL .{ MOTILITY WITH A NOVEL ANTRAL WALL MOTION DETECTOR r4 .,.: 8.3.1 Introduction Low amplitude contractions of the gastrointestinal wall, which do not occlude lhe lumen, may produce little or no change in intraluminal pressure (Carlson et al 1966; Christensen 1987) and will therefore not be recognized by intraluminal manomelry (Chapter 21. Carlson et al (1966) demonstrated radiologically that some antral contractions deform the antral wall but do not occlude the lumen. These observations have important implicalions, since recent advances in knowledge about the role and function of the human pylorus depend on the correct interpretation of not only pyloric, but also of antral pressure recordings.

Recently, prolonged pyloric manometry using the sleeve/side-hole techniques described in Chapter 2, have identified apparently isolated pyloric pressure waves, or lPPWs, during fed and fasting motility, and an increased incidence of IPPWs after ingestion of nutrients and intraduodenal infusion of nutrients or acid (see Chapter 4, and especially Heddle et al 1988bac, Houghton et al 1988aab). A requirement for the manometric definition of isolated pyloric pressure waves in these studies was that there be complete suppression of associated distal antral and proximal duodenal prêssure waves. ln the light of Carlson's fluoroscopic observations, I and having regard for the limitations of intraluminal manometry (Chapter 2), there remains some uncertainty about whether manometrically defined isolaled pyloric pressure waves are generaled by truly localized contractions, or whether they could be due to less localized events. Such r 83

events could beg¡n as low amplitude contractions in the proximal antrum and propagate aborally, producing a narrow zone of lumen occlusion recorded manometrically only at the pylorus. This ¿i latter possibility is made especially plausible by the fact that many IPPWs normally occur at approximalely the antral frequency (Chapter 5; Heddle et al 1988b). Resolution of this uncertainty is desirable since there is increasing evidence that localized pyloric contractions occur during suppression of antral contractions, retard transpyloric flow, and may be important in the regulation of gastric emptying (Chapter 6). Furthermore, understanding of the normal patterns of antropyloric molility, and therefore of the processes of triturition and gastric emptying, may be enhanced by more sensitive recordings of antral contractions of all amplitudes.

The aim of this set of experimenls was to evaluate a new recording device designed lo detect those contractions which substantially narrow the antral lumen, but which may not result in lumen-occlusion. The sensitivity of this antral wall movement detector, ortransducer, was compared with that of conventional side-hole manometry under three experimental conditions. This device was then used to examine the question of whether IPPWs are associated with suppression of all movement of the antral wall within the limils of intraluminal detection, and so can truly be considered to be due to significant contraction of only an isolated zone at the pylorus.

l:1

l[J 8.3.2 Materlals and Methods i 8.3.2.1 Subjects Studies were performed on 7 healthy volunleers (4 men, 3 women) aged 20 to 29 years (mean age 2'l years).

8.3.2.2 Experimental Protocol Pyloric, antral and duodenal motility were recorded for 45 minutes under each of 3 conditions:

1 ) fasting 2) after ingestion of a 100 gram ground beef burger, identical to lhe standard meal used in gastric emptying studies (Chapter 7), but without the isotopically-labelled chicken liver. The meal was consumed, without liquid, over 5 minutes. 3) during and after an intraduodenal infusion ol 259. dextrose, delivered at a rate of 4 ml/min, designed 1o stimulale isolated pyloric pressure waves (Chapler 5).

8.3.2.3 Measurement of Motility $ I a) Antral wall movement The intraluminal antral wall movement transducer was designed and manufactured in the ; Departments of Experimental Surgery at the Academic Hospitals in Utrecht and Rotterdam, The Netherlands. lts design was based on methodology used for in vivo measurements of intra-

I 84

arterial diameters in animals (Van der Schee and Geldof 1985), and its detailed construction is described in that paper. A similar lransducer was used in two previous preliminary studies of human antral motility in post-surgical patients (Van der Schee et al 1981; Looijen et al 1987). ln the present series of experiments, lhe transducer consisted ol a 2.5 cm long wire ellipse wilh a lransverse diameter of 1.5 cm, which was fixed at ils proximal end to the catheter tubing (Figure 8.3). The distal end of the ellipse was connected to a central rod which passed within lwo parallel conducting coils through which an electrical current was passed (10 kHz 2mA). Compression of the elliptical wire spring, which required a minimum force of 0.02 N, caused movement of the central rod within these coils. The changing electrical signal which resulted was processed by an amplifier/ demodulator, and printed out continuously on a chart recorder (Figure 8.4). The transducer was calibrated at the beginning of each study so that reduction of the spring diameter to 5 mm resulted in a full scale deflection of the pen on the chart recorder (Figure 8.5). Calibrations were rechecked at the end of each study.

b) Manometry The manometric recording assembly differed slightly from that described in Chapter 7 (Figure 8.3). lt included a 4 cm long sleeve sensor to measure pyloric motility. Three side-holes were located in the antrum, one at each end of the elliptical spring 4.5 cm and 2.0 cm proximal to the sleeve (A1 and A2 respectively), and one at the most orad end of the sleeve (A3). Two further T iu side-holes were used lo measure duodenal motility, and were located at lhe most distal end of the sleeve (D1) and 2 cm further distally (D2).

8.3.3 Data Analysls 8.3.3.1 Analysis of Records Deflections in antral transducer tracings of at least 4% lull scale excursion could be easily identified on the tracings, and were scored as antral contractions. Bench testing indicated that this deflection was equivalent to a 0.2 mm compression of the elliptical spring. A transducer deflection was considered to be associated with a pressure wave if its onset occurred within 5 sec of the sharp initial upslroke of a pressure wave in either of the two anlral side-holes at each end of the wire ellipse (41, A2), or within 15 sec of the sharp initial upstroke of a wave recorded by the most distal antral side-hole (A3) or the sleeve. These time intervals were judged to be appropriate based on preliminary observations of fed antral motility using this assembly. Antral pressure waves recorded in different side-holes were said to be associated with each other if the onsets of their sharp initial upstrokes occurred within 5 s (see Chapter 7), t I Pressure waves recorded by the sleeve were scored when they were at least 10 mmHg in ; amplitude and, as in other studies, were classified as isolated pyloric pressure waves (lPPWs) if there was an absence of associated pressure waves of any magnitude in the adjacent antral or

! 85

SLEEVE

TMPD TMPD

SIDE HOLE

ROD

rl SPRING colLs wEt HT CATHETER TUBING

TO SLEEVE

lcm WALL MOTION DEVICE

Flgure 8.3 Schematic r€pr€ssntation of the recording assembly in position with the sleeve located across lhe pylorus. t The side-holes at either end of lhe sleeve are used for transmucosal potential difference (TMPD) I measurements. Detail of the antral wall motion delector (antral transducer) is shown in the enlargement. l

t B6

Ghart Recorder

Ampllf ierl ^.- Ge nerator Demodulator ^ ^-^. Â r. nrr tn out o o ¡ïr À^ o o o

Transd ucer ( Side-holes Sleeve ,

Flgure 8.4 Schematic representation cjf the equipment used to record anlral wall motion and pyloric motility. The constant electrical output from the generator was altered by compression of the transducer ellipse. The resulling signal was processed by the amplifier/demodulator and printed out continuously on the chart recorder, simultaneous with pressure tracings from the sleeve and the side-holes in the antrum and duodenum.

100

% DEFLECTION

o 7 1o89111012

INTERNAL BING DIAMETER (mm)

Flgure 8.5 Example of transducer calibration. At the beginning of each study, prior to intubation, lhe lransducer was drawn in random order through a series of plastic cylinders with inner diameters of 5-12 mm. There is an approximately linear relationship belween the size of lhe transducer deflection and the degree of compression of the wire ellipse. 87 duodenal TMPD side-holes. ln previous sleeve/side-hole studies of pyloric molility (for example, Heddle et al 1988a,bac; Houghton et al 1988aab) and in the other studies described in this lhesis, only anlral pressure changes which were grealer than 10 mmHg in amplitude were scored as antral contractions. However, in order to maximize the sensitivity of the side-hole pressure recordings in the present sludy, pressure waves recorded by the side-holes in the antrum were scored if they were easily discernible as an increase in pressure of at least 2 mmHg above normal baseline respiratory fluctuations. To help judge most appropriately whether an antral pressure wave of <10 mmHg in amplitude should be scored, its relationship with other pressure waves and transducer deflections, and its liming with respect to the known antral slow wave frequency of 3/min, were taken into account.

8.3.3.2 Statistical Methods The number of pressure waves and transducer deflections were counted for each 5 minute period and expressed as lhe mean number of pressure waves/min. Results from the different sensors and recording sites were compared with the Wilcoxon matched pairs signed-rank test. As with other studies, a p value of <0.05 was considered significant.

8.3.4 Results 8.3.4.1 Subject Tolerance and Sensor Positioning The nasogastric intubation, ingestion of the lest meal, and intraduodenal dextrose infusions were well tolerated in all subjects. The position of the calheter across the pylorus was difficult to maintain during lasting and after ingestion of the meal, although displacement of the catheter could be readily detected from the manometric patterns and by a change in TMPD measuremenls. ln two studies there was difficulty in achieving stable location of the sleeve across the pylorus until after the ingestion of the test meal, and two other studies ended prematurely because of proximal displacement of the tube. Consequently the experiments yielded satisfactory recordings under the following conditions: fasting in 5 subjects; after the solid meal in 6 subjects; following intraduodenal infusion of dextrose in 5 subjects. During these recordings, the manometric assembly was correctly located across the gastroduodenal junction lor 86i" of recording time (569 out of 660 minutes).

8.3.4.2 Antral Motility The numbers of events recorded by the antral transducer and antral side-holes, under the three study conditions and throughout the entire study, are displayed in Table 8.3 and Figure 8.6.

(i) Fasting: Contractions were recorded during phase ll of the interdigestive migrating motor cycle in 5 subjects, and during phase lll in one subject. A lotal of 104 antral transducer deflections was recorded. ln all but one subject, the antral transducer recorded more deflections than any of the three antral side-holes. ln each subject, the more proximal side-holes recorded 88

FASTING MEAL INTRADUODENAL DEXTROSE no. to No. olto No. o/o

.TRANSDUCER * * 104 100 423+ 100 124 100 A1 19 18 19 4 0 0 A2 28 27 184 43 11 9 A3 46 44 280 66 61 49

Transducer Only 69 175 78 Side-hole Only 20 12

Table 8.3 Number of events recorded in all subjects by the antral transducer and the three antral side-holes situated at 4.5, 2.O and 0 cm above the sleeve (41 , 42, Ag respectively). Percentages are referenced lo total transducer-recorded events. Many contractions were recorded only by the transducer ('transducer only'); few events were rscorded by the side-holes without a lransducer deflection ("side-hole only'). *p<0.05 cf A1, A2, A3. +p<0.05 cf 41, A2.

consistently fewer pressure waves than those located distally. Of the 104 lransducer deflections, 69 (66%) wsre not associated with any change in side-hole pressures. Only 7 (17%l of the antral side-hole pressure waves occurred w¡thout any associated transducer deflection, and 6 of these 7 waves were recorded by the most distal antral side-hole.

(ii) Fed: Antropyloroduodenal pressure waves were the dominant manometric pattern after ingestion of the solid meal. When high amplitude pressure waves were recorded by the proximal antral side-holes, they were usually associated with correspondingly large deflections in the transducer tracing (Figure 8.7a). At times, regular transducer deflections were recorded at an intrinsic frequency of up to 3.0/min when pressure waves were recorded only from the most distal antral side-hole (Figure 8.7b), or when there was no detectable pressure rise in any antral side-hole (Figure 8.7c). Of 428 transducer deflections, 175 (41%) were not associated with any side-hole pressure wave. 7"/" of pressure waves occurred without any transducer deflection.

(ii¡) lntraduodenal dextrose: The suppression of antral motility (p<0.05) produced by intraduodenal dextrose (Figures 8.8,8.10 and 8.11) resulted in only 124 lransducer deflections being recorded in 210 minutes of recording after the start of the dextrose infusion. ln the one 89

l OOo/" 700

600

500 Number of 59o/o Recorded 400 Evcnts 300 3 4o/o

200

o/o 100 6 : TRANSDUCER A3 A2 A1

Flgure 8.6 Total numbers of events recorded by the antral transducer and antral side-holes under all study conditions 90

A B c f

TR DUGER /\-J\-- -/.\^ I ^

mmHg ¡lO

1 o sô¿êôôÂÁ4 À.ô.ô-ô.êô¿ô

A2 '"1 ^^*^* æâ¿ â.4,Âáââ4.ô,

3 '"J AêÂô¿4áâ

40 LEEVE q¡¡Ê\./\-,/- o

D1 èâ.444.âs

30 aeca

Flgure 8.7 Patterns of pressure waves observed during 3/min transducer deflections, recorded after ingestion of the solid meal. 91

DI.gTAL ANTRAL .J SIDEHOLE m¡nHg

TRANSDUCER % Dollectlon

3() 3ac¡

Flgure 8.8 Contlnuous segmsnt of antral transducer tracing for 5 min before (upper pair of tracings) and 5 min after (lower pair of tracings) the start of the lntraduodenal lnfusion of 25/" dextrose (hatched bar). 92 subject who had no IPPW response, 80 transducer deflections occurred throughout the 45 minute period after dextrose. ln three subjects, transducer deflections continued after the start of the dextrose infusion, but then disappeared within 5 minutes (Figure 8.8), before the appearance of pressure waves classified manometrically as lPPWs. The mean amplitude of transducer deflections in the 5 minutes after dextrose was less than in the 5 minutes before dexlrose (6.1 vs 9.3 mm). All other transducer deflections recorded in the 45 minutes after the start of the intraduodenal dextrose infusion occurred at least 20 minutes after the end of the dextrose infusion, when the number of IPPWs had diminished, and with only lwo exceptions were not associated with IPPWs (Figure 8.10).

8.3.4.3 lsolated Pyloric Pressure Waves Consistent with previous studies (Chapter 4), the infusion of dextrose into the duodenum was associated with an increase in the number of IPPWs in 4 of 5 subjects (p<0.05, Figure 8.10). There were 42 sleeve-recorded pressure waves which satisfied the manometric criteria for IPPWs during fasting,52 after ingestion of the solid meal, and 181 during and after the intraduodenal infusion of dextrose (Table 8.4). After intraduodenal dextrose, only 2 IPPWs (1.19.1 wsre associated with any transducer deflection, and transducer tracings otherwise remained flat (Figure 8.11). Similarly, there was no transducer deflection with any of the IPPWs recorded during fasling. However,9 out of the 52 IPPWs (18%) recorded after the solid meal wêre associated with significant deflection of the antral transducer. After the hamburger, sleeve-rscorded pressure wavês werê occasionally observed in the absence of pressure waves in the most distal antral side-hole, but with associated pressure waves, usually <1OmmHg in amplitude, occurring in the more proximal antral side-holes (Figure 8.9). These sleeve-recorded waves were classified as IPPWs according to previous criteria (see above and Section 7.5.1). Of the 9 IPPWs after the meal which were associated with transducer deflections, 7 were also associated with these low amplitude pressure waves in lhe proximal anlrum.

FASTING MEAL INTRADUODENAL DEXTBOSE

IPPWs 42 52 181

IPPWs plus 0 9 (18%) 2 (1.1%l Transd ucer

Table 8.4 The relationship between the total numbers of IPPWs under the differenl study condilions, to the number of transducer-recorded evenls. Few IPPWs were temporally associated with a lransducer deflection (.lPPWs plus Transducer"). 93

% DEFLECTION '":l TRANSDUCER

mmHg A1 l .J A2

A3 l

'.1 a a SLEEVE a

D2 .1 30 ¡oc

Flgure 8.9 Tracing after ingestion of the solid meal demonstrating three sleeve-recorded pressure waves classified initially as lPPWs, associated in time with transducer deflections (asterisk). Note, however, that associated pressure waves are presenl in a proximal antral side-hole (42), despite an absence of any pressure waves in the most distal antral side-hole (A3). 94

25o/" Dextrose 64 Antral transducer deflections H rppws 3.0

2.0 Recorded Events No./min 1.0

-20 -10 0 10 20 30 40 50 TIME minutes

Flgure 8.10 Relationship betwesn sleeve-recorded IPPWs and antral activity detected by the transducer, demonstrated in a graph from one subject. During maximal IPPW stimulation, no antral activity is apparent. A rise in antral transducer deflections occurs before the end of the IPPW response, however analysis of the tracing indicated that these events wsre not associated wilh lPPWs. 95

roo

TRANSDUCER % o

mmHg .to A1 o 40

^2 o

A3 '"1

SLEEYE

40 Df o

"l 30 Êeca

Flgure 8.11 lsolated pyloric pressurs waves recorded by the sleeve during intraduodenal infusion of 25/" dextrose. Note that no pressure waves are present in the anlrum or duodenum, and the antral transducer tracing indicales the lransducer is relaxed; there is no discernible evidence of antral wall movement during the dextrose infusion. 96

8.3.5 Summary of Results The hypofhesis fhat isolated pyloric pressure waves in humans occur in the absence of even low amplitude antral contractions, was examined with sleeve/side-hole manometry and an antral wall movement deteclor in 7 healthy volunteers. Antral contractions were detected by the wall motion detector with greater sensitivity than antral side-holes, possibly reflecting lhe occurence of non lumen-occluding antral contractions. Overall, only 51 lo of anlral transducer deflections were associated with a change in antral side-hole pressures. Eighty-nine percent of antral side- hoie pressure waves were associated with an indication of antral wall motion. Of the pressure waves recorded by the sleeve which were classified as isolated pyloric pressure waves, none was associated with antral transducer deflection during fasting, 1.1"/" allør intraduodenal dextrose, and 18/" after the solid meal.

8.4 DISCUSSION These studies investigated technical aspects of intraluminal manometry, which have important implications for fhe correct interpretation of pyloric motility recordings. The results suggest that transpyloric intubation with a sleeve assembly can be performed without causing undue alterations in pyloric motility. However, it is clear that the limitations of sleeve/side-hole manometry must be recognized and considered in the design of manometric assemblies and the interpretation of pressure recordings then obtained.

There have been no previous experiments which have compared the fed patterns of anlral, pyloric and duodenal motility during transpyloric and terminal ileal intubation, with manometric assemblies of essentially identical diameter and with a test meal of similar composition. The results of the present study confirm that ileal (Read et al 1983), but not transpyloric (Muller- Lissner et al 1982; Longstreth et al 1975) intubation has a substantial effect on gastric emptying and antral motility. lt has previously been proposed that the rate of gastric emptying of digestible solid food is primarily determined by antral contractions (Camilleri et al 1985; Meyer 1g87), and the resulfs of the present study support the existence of such a relationship. However, fhere is considerable evidence that changes in pyloric motility, as well as that of the duodenum and proximal stomach, are also of importance in controlling gastric emptying (Tougas et al 1987; Houghton et al 1988aab; Heddle et al 1988a,b&c; Azpiroz and Malagelada 1988; for a detailed discussion see Chapter 6 and the review by Meyer 1987). The similar rate of gastric emptying in the non-intubated and duodenal intubated groups supports the contention that gastroduodenal motility is not significantly altered by transpyloric intubalion with a short and narrow cafheter. Although this is reassuring, it remains only a reasonable assumption. One could postulate, for example, that reductions in both antral and pyloric motility could occur with intubalion, with no net effect on the rate of gastric emptying. This does not seem likely, since 97 antral pressure waves are oflen recorded by transpyloric manometry at the maximal rate of approximafely 3/min. However, manometric information about the amplitude of antral wall contractions is incomplete (Chapters 2 and 8.3). The only way of finally resolving this issue would be to record pyloric and antral contractions with a third, non-invasive technique, simultaneous with gastric emptying and manometry, in intubated and non-intubated groups. At prosent, apart from manometry, there is no technique which is suitable as a "gold standard' for measuring pyloric motility in humans. Possibly, however, this may be achieved to some degree with sophisticated ultrasonography (Chapter 2), although the presence of a catheter would be likely to cause lechnical problems. Future improvements in the spatial and temporal resolution of radionuclide methods, with the use of very rapid frame-rates, could well provide a suitable alternative (Chapter 2). ln studies which involve distal small intestinal intubation, unless intubation is completed overnight, the long lime required to position the manometric assembly will result a late afternoon starting time for measurements. This was the case in the group requiring distal ilêal inlubation in the study reported in this chapter. Goo et al (1987) recently demonstrated a circadian variation in gastric emptying, with gastric emptying of a solid meal in normal male volunteers being more rapid at 8 a.m. than at I p.m.. This suggests that the association in the present study between ileal intubation and both slower emptying of the solid meal and reduced antral contractions, could possibly have been due to differences in the time of the day at which the test was performed, and the longer period of fasting in the ileal intubation group, rather than ileal intubation per se. The difference in gastric emptying rates observed by Goo et al (1987)' however, was substantially smaller (the mean retention of a 150g meat meal at 60 minutes was present approximately 5217" at I a.m., compared lo 64T" at I p.m.) than was recorded in the study. Furthermore, Read et al (1983), in their study of ileal intubation, gave both intubated and non-intubated subjects the test meal at between 9 and 10 a.m. and still found substantially slower gastric empty¡ng in the ileum-intubated subjects. ln the Present study, therefore, the inverse correlation observed between the time for intubation and the frequency of antral waves is unlikely to represent a causal relationship. There is no information about whelher the duration of fasting affects emptying of subsequent meals, but this also appears an unlikely explanation of the results.

Other studies have demonstrated that gastroduodenal motility may be modified by stimuli distant from the stomach and duodenum, including painless rectal distension (Youle and Read 1984)' stress caused by cold pain and labyrinthine stimulation (Stanghellini et al 1983; Thompson et al 1982) and terminal ileal lipid infusion (Read et al 1984). Gaslric emptying is delayed by stressors, such as labyrinthine stimulation and cold pain (Stanghellini et al 1983; Thompson et al 19g2). Psychological stress, such as that induced by a dichotomous listening test, inhibits fasting 98 human jejunal molor activity (McRae et al 1982), but its effect on gastric emptying appears 1o be inconsistent (Cann et al 1983; O'Brien et al 1987; Chapter 10). Nevertheless, it is possible that the inhibition of gastric emptying in the ileal intubation group was caused by a nonspecific mechanism such as the induction of stress or nausea, due to the longer time taken for each experiment, although the latter symptom was not reported by any volunteer. The effects of ileal intubation may also have been mediated by mechanical stimulation of the mucosa of the small intestine which has been demonstrated to inhibit gastric contractions in the dog (Brunemeir and Carlson 1915). Possibly, such an effect could be mediated by vagal mechanoreceptors responsive to stretch and distension, as described in Section 5.4.2. Although mechanical stimulation of the stomach, oesophagus and proximal small intestine arising from the presence of a manometric assembly apparently does not modify gastric motility or gastric emptying, a much larger number of mucosal receptors may have been stimulated in the ileal intubation group. The recenl paper by Lin et al (1989) demonstrating that inhibition of gastr¡c emptying of glucose is dependent on the length of small intestine exposed to glucose is consistent with this hypothesis (Section 5.4.2 and Chapler 9).

It is known lhat a proportion of antral contractions does not result in any intraluminal pressure change, but the sensitivity of manometry for recognit¡on of antral contractions is not well defined. ln particular, no previous study has evaluated human pyloric or antral motility using informalion from an intraluminal antral wall motion detector. The wall motion detector employed in the sludy reported in lhis chapter was demonstrated to be substantially more sensitive lhan even the most distal of the antral side-holes. The results indicate that there is no detectable movement of the antral wall in association with isolated pyloric pressure waves (lPPWs) during fasting or after stimulation by intraduodenal dextrose infusion. These findings support the postulate that the manometric pattern of IPPWs is produced by very localized contractions occurring at the pylorus. On the other hand, recordings from the antral wall movement detector indicated that a proportion of IPPWs recorded after food were apparently less localized, suggesting that existing criteria may lead to significant mis-classification of pressure waves at the pylorus and/or distal antrum under some circumstances.

Antral contractions will be detected with the greatest senstivity when the sensor is positioned in the narrowest luminal segment, that is in the most distal antrum. The antral transducer was positioned very close to lhe pylorus with the aid of TMPD recording. Because of the lenglh of lhe sleêve, however, recording assembly movements ol 2-3 cm may have occurred without a change in TMPD. Thus it was possible for the transducer to drift slightly orad from the extreme distal antrum, and so contractions which substantially narrowed only the terminal antral lumen may not have been detected. The extent to which this may have led to under-recording of antral contractions in the present study cannot be determined, however it is unlikely to have occurred oo frequently, since antral contractions were mostly detected at the maximal antral frequency. Some antral contractions may have failèd to causs transducer deflection even when lhe sensor was accurately localed in the very distal antrum, if the lumen was not narrowed to the maximum lransverse diameter of the transducer (1.5 cm).

The relative timing and association with antral pressure waves suggests that deflections of lhe transducer wers almost certainly due to antral contractions. Conceivably, lateral forces on the assembly unrelated to antral contraclion, for example external antral compression by adjacent organs such as lhe colon, could also produce a signal from the wall motion detector. However, since the deflections wers recorded at a regular antral rate of approximately 3/min, lhe incidence of 'false positive' deflections was probably low.

The finding of greater sensitivity of the antral wall motion detector compared to intraluminal side-holes is consistent with a previous study reported in an abstract (Van der Schee and Geldof 1985). This discrepancy between recording techniques cannot be reasonably explained by the transducer preventing closure of the antral wall over the side-holes during a contraction, since only a very small force (0.02 N) was required to fully compress the transducer, and the side- holes were located perpendicular to the plane of the transducer whereas antral lumen occlusion occurs in three dimensions. Thus, the discrepancy is likely to be due to the occurrence of low amplitude antral contractions which deformed the antral wall, but did not produce lumen occlusion or any change in intraluminal pressure. The more proximal side-holes recorded relatively few contractions, consistent with radiological observations that many aborally propagating antral waves begin proximally as non lumen-occluding low-amplitude contractions, which increase in amplitude and occlude the lumen as they pass distally towards the narrow gastroduodenal junction (Carlson et al 1966).

Recordings with the antral probe provide new information about the patterns of pyloric contraction. Previous studies using intraluminal tonometers (Thomas et al 1934), perfused catheters and balloons (Carlson and Litt 1924; Thomas et al 1934; Quigley et al 1942) suggested thal under some circumstances the pylorus could contract independently, and that this may be associated with complete suppression of antral motor activity. These studies were methodologically inadequate for accurate pyloric manomelry (Chapter 2), however, localized pyloric contraction was also reported during early fluoroscopic observations (Thomas et al 1934). Later scepticism about the existence of this pattern of motility arose f rom inconsistencies amongst studies, and uncerlainties about the technical requirements for pyloric manometry. As discussed in Chapter 4, rscent studies using sleeve/side-hole manometric assemblies and transmucosal potential difference localization have now convincingly demonstrated the occurrence of isolated pressure waves at the gastroduodenal junction which 100 were associated with absent distal antral pressure waves (Heddle et al 1988b&c; Houghton et al 198gaab; see also Chapters 9 and 1O). Nevertheless, the occurrence of low amplitude non lumen-occluding antral contractions which do not produce any change in antral pressure (Carlson et al 1966) raised the possibility that some manometrically defined IPPWs could represent the final lumen-occluding compononl of a longer conlraction which began as an undetecled ripple in the proximal antrum. The present results indicate that during fasting and after stimulation by intraduodenal dextrose, IPPWs do indeed appear to be due to very localized pyloric conlraction. After ingestion of the hamburger, however, 181" of waves classified as IPPWs according lo previous criteria were associated with lransducer deflection. Heddle et al (1988b) observed that IPPWs induced by intraduodenal lipid infusions occurred over a very narrow manometric zone, with 92% being less than 9mm wide (Chapters 2 and 4). This finding is supported by recent radiological observations reported anecdotally and in abstract (White et al 1983; Tougas et al 1987). The anatomical basis for this phenomenon is discussed in Chapter 1. ln the present study, pyloric motility was recorded with a sleeve sensor which was 4cm long. The sleeve sensor is particularly suited to the demands of pyloric manometry, but since its length leads to protrusion from either side of the distal pyloric muscle loop, it will also record pressure changes in the immediately adjacent antrum and duodenum (Chapter 2). This recording limitation has been recently addressed by the inclusion of side-holes spaced at 1cm inlervals along the length of the sleeve, to differentiate less localized waves from IPPWs more effectively (Heddle et al 1988c). Such a series of side-holes was also employed in the other studies reported in this thesis. Using this approach, analysis of recordings after ingestion of a solid meal (Chapter 9) showed that 13% of sleeve-recorded waves which were not recorded in the antral or duodenal TMPD side-holes, occurred over a luminal length of lhe sleeve more consistent with these waves being short antropyloric or pyloroduodenal events, rather than the narrow IPPWs reported by Heddle et al (1988b). Possibly, the sleeve-recorded pressure waves classified as 'lPPWs', but which were associated with a transducer deflection after food in the present study, could have represented similar, less localized waves. This indicates the need to include side-holes along the length of the sleeve in order to classify sleeve-recorded waves as accurately as possible. Motility patterns may, however, be different after ingestion of meals of different volume or viscosity, for example after a large liquid meal, and lhe relationship of pyloric pressure waves to antral wall movement under such conditions is worthy of separate evaluation.

lntraduodenal infusions of dextrose induce dose-related stimulation of IPPWs and suppress antral and proximal duodenal prossure waves (Heddle et al 1988c; Chapler 41. ln the present study, the gradual diminution in amplitude of transducer deflections after intraduodenal dextrose, prior to the appearance of lPPWs, suggests a gradation of suppression of antral motility and activation of localized pyloric contractions. Such flexibility of response would provide controlled braking of 101

¿\ gastric emptying, via minute-to-minute variations of propulsive antral pumping (Kelly 1980) and retardation of transpyloric flow by isolated pyloric contraction (Tougas et al 1987). The characterislic pattern of gastroduodenal motility reportod after duodenal receptor stimulation by intraduodenal infusion of high calorie nutrient solutions (Heddle et al 1988b4c, and see also Chapler 4) may repressnt an extreme end of this spectrum.

The results of this study have implications for the interpretation of sleeve/side-hole manometric recordings from lhe gastroduodenal junction in which information from a wall motion detector is not available. Firstly, since some very low amplitude anlral pressure waves in the present study were associated with deflections of the antral wall motion detector, and clearly represented antral contractions, the sensitivity of manometric recordings would probably be improved if antral pressure waves were counted irrespective of their amplitude, as long as they were easily discernible above the normal baseline and were not due to straining. Secondly, because some antral pressure waves may be recorded proximally despite no detectable pressure rise in the most distal antral side-hole, sleeve-recorded wavês should only be classified as IPPWs if there is an absence of pressure waves in any antral side-hole, not just in the antral TMPD side-hole as previously defined (Heddle et al 1988a,bac; Houghton et al 1988aab). This means that recording assemblies for detailed pyloric manomelry must include a number of antral side-holes, as well as side-holes spaced at <1cm intervals along the length of the sleeve sensor. This feature was incorporated into the assemblies used in the other studies reported in this thesis.

The marked difference between the number of pressure waves recorded by the different anlral side-holes indicates that precise and preferably continuous feedback of assembly location is essential for proper interpretation of antral manometry. The use of antral and duodenal TMPD measurements from either end of the sleeve sensor during all studies of antropyloroduodenal motility (Houghton et al 1988aab; Heddle et al 1988bac; Chapters 9 and 10) is probably the best available technique, since it continuously indicates assembly position to within 2-3 cm. Adjustment of assembly position at regular intervals is necessary to maintain the sleeve accurately across the gastroduodenal junction according to TMPD criteria. lf assembly position could be controlled even more precisely, there would probably be worlhwhile gains given the substantial differences in pressure patterns over short luminal distances in the distal antrum. Achievement of such control of assembly position presents a major technical challenge.

Few studies have examined the sensitivity of antral side-hole manometry by direct comparison with other lechniques. Carlson used cineradiography and antral balloon manomelry, and observed the occurrence of antral contractions without associated pressure signals (Carlson et al 1966). You and Chey (1984) compared the performance of antral side-holes with serosal electrodes in humans after bethanechol stimulation, and found that the number of pressure waves recorded by 102

the side-holes was <50% of the electrical patterns expected to result in a conlraclion. However, the rolative positions of the electrodes and side-holes in that study were not precisely correlated. Valori et al (1986) found that perfused side-hole manometry in anaesthetized dogs recorded up lo 98% of the field or vagally stimulated antral contractions indicated by serosal strain gauges. lntraluminal sensors described as strain gauges were less sensilive, but insufficient detail was given about the design and validation of these sensors to judge whether they recorded movement, deformity or prêssure. lt is not clear whether the comparable sensitivities of the serosal strain gauges and manometry in that study was due 1o the study conditions, or to the attention paid to locating side-holes and strain gauges at exactly the same site. Differences in sensitivity may well have been found if relatively low intensity electrical field stimulafion had been used. The difference between the studies reported in this chapter and those of Valori et al (1986) is likely to be explained by the different testing approach, lhat of Valori being unrePresentative of the normal spectrum of motor activity.

8.5 CONCLUSION These studies have defined more precisely the best approaches to the use of combined sleeve/side-hole manometry for the evaluation of pyloric manometry. Both the limitations and advantages of these techniques must be recognized for oplimal interpretation of pressure recordings from this region. Then, used appropriately, manometry alone should generally allow q accurate judgements to be made regarding patterns of pyloric motility, although lhe potential for intubation-induced alteralions in antral motility must be recognized, especially if the recording catheter extends into the distal small bowel, or the duration of the study is very long. Despite fhe increased experimental complexity involved, the combination of manometry with an antral wall motion detector may be useful for research purposes, given the additional information such a sensor would provide. Methodological insights gained from the use of an antral wall motion

I detector in this study have furthered understanding of the patterns of coordinated motility displayed by the pylorus, and the potential role of the pylorus in the physiological control of gastric emptying.

þ 103

CHAPTER 9

Control of Pyloric Motility by Stimulation of Small Intestinal Receptors

9.1 INTRODUCTION The stimuli which exert feedback on the rate cf gastric emptying have been well characterized, but the effect of many of lhese stimuli on pyloric motility has not been eslablished, despite the '\ potential role of the pylorus in the contol of gastric emptying. Although there is now considerable tl 1Ì information about the influence on the pylorus of upper small intestinal nutrients, such as fat and carbohydrate (Tougas et al 1987; Heddle et al 1988b&c; Chapter 5), the motor mechanisms by which stimulation of the distal small bowel by nutrients delays gastric emptying (Chapter 5) are poorly understood. Furthermore, although some of the neural and hormonal pathways which mediate such responses in animals have been investigated (Reynolds and Ouyang 1985; Allescher et al 1988), there have been no adequate studies of the pharmacology of these responses in humans. The results of experiments which examine the neurochemical mechanisms mediating the human pyloric motor response to intraduodenal dextrose, and the effect of infusion of triglyceride into the distal small intestine on patterns of pyloric motility and gastric emptying, are described in this chapter.

9.2 A STUDY OF THE ROLE OF MUSCARINIC MECHANISMS IN THE PYLORIC MOTOR RESPONSE TO INTRADUODENAL DEXTROSE IN HUMANS 9.2.1 Background Studies in the dog indicate that the canine pyloric motor responses 1o duodenal receptor t I stimulation by acid may be mediated by orally projecting cholinergic excitalory fibres within the duodenal wall (Allescher et 1987; Bovell Chapter This may not be the case in ; al et al 1987; 5). the cat, where atropine was found to have no effect on the pyloric motor responses after intraduodenal acid or amino acid infusions (Reynolds et al 1984 & 1985), suggesting possible

t 104

interspecies variations. Several previous studies in humans, which have examined the effect of atropine and naloxone on lhe pyloric molor response to duodenal acidificalion (Valenzuela et al 1976; Dooley et al 1985), have been lechnically inadequate (see Section 9.4). Another study, using a more reliable manometric approach, found that naloxone had no effect on the pyloric motor rosponse lo intraduodenal lipid infusion (Bovell et al 1987), although opiates probably play

a role in controlling other aspects of upper gastrointestinal motility in humans (Camilleri et al

1986). Because of the paucity of informalion about the pathways controlling pyloric motility in humans, it remains unclear whether the human pylorus is pharmacologically similar to the dog or the cat models in which most investigations have been performed. The aim of this study was to investigate whether the pyloric motor responses to intraduodenal dextrose infusions in humans are mediated by atropine sensitive pathways.

9.2.2 Methods 9.2.2.1 Subjects

Studies were performed on 4 men and 6 women, aged 19-21 years and weighing 47-78kg.

9.2.2.2 Experimental Design

The manometric assembly was positioned across the pylorus as describ€d (Chapter 7). Through each experiment, three intraduodenal infusions oÍ 25"/" dextrose were given at a rate of 4 .T ml/min, delivering 4.0 kcal/min into the duodenum (Figure 9.1). The first infusion was continued '1f 'rÈ for 15 minutes after it had produced an established phasic pyloric motor response, defined as the occurrence of IPPWs at a rate of at least 0.8/min for 5 minutes (Heddle et al 1988c). The second and third dextrose infusions were given for the same length of time as the first infusion. Each dextrose infusion was preceded by a 10 minute control period, which was only started when motor patterns had returned to a basal state. During the second dextrose infusion, inlravenous atropine was administered 5 minutes after the commencsment of the pyloric phasic response. Atropine was given as a bolus of 15 ug/kg, followed by a maintainance infusion of 4 ug/kg/hour, which was continued until the end of each study. Consequently, the third dextrose infusion was given during what was judged to be steady-state atropinization, and recordings were continued until at least 15 minutes after the end of this infusion.

Heart rate was monitored by palpation of the radial arterial pulse, 5 minutes before and afler the atropine bolus and at 15 minute intervals thereafter. At the same time interuals, each volunteer

was asked to report the degree of mouth dryness experienced, using a visual analogue scale of 0 to 10, with 0 being 'normal, nol dry' and 10 being 'as dry as possible'. Volunteers were examined for pupil dilatation and asked about blurred vision. They were also asked lo report any i olher unusual symptoms at any time.

r 105

Atroolne Bofus l5mcg/kg lnfuslon 4 mcg/ kg/h I

Intraduodenal Dext rose u777777t v77777Ã 25o/o 4ml/mln

0 20 40 60 80 1 00 120 140 d 'j TIME minutes j

Flgure 9.1 Schematic reprssentation of the experimental protocol. The three identical intraduodenal dextrose infusions (hatched bars) were each preceded by a 10 minute control period. lntravenous atropin¡zation continued until the end of each study, and manometry was performed throughout. Heart rate, mouth dryness and mydriasis were monltored regularly throughout the study.

I 106

9.2.2.3 Pressure measurement

The manometric assembly and the recording techniques were similar to those described in Chapter 7. ln addition to the antral and duodenal TMPD side-holes, two further antral side-holes were situated 2 and 4 cm oral 1o the sleeve, and three addilional duodenal side-holes recorded duodenal pressures 3,9 and 12 cm aboral to the sleeve.

9.2.2.4 Data Analysis

Records were divided into 5 minute periods for 30 minutes after the time of arrival of lhe dextrose infusions in the duodenum, and for lhe preceding 10 min control periods. With each dextrose infusion, the mean number of IPPWs (lPPWs/min) and the mean basal pyloric prêssure (mmHg) were delermined for each 5 min period. Analysis of the initial effect of atropine during the second dextrose infusion was referenced to the time of administration of the atropine bolus, and pressure recordings in the 10 minutes immediately after the atropine bolus were compared with the 10 minutes immediately before. Scoring of pressure waves and determination of basal pyloric pressuro were performed as outlined in Chapter 7. The duralion of the IPPW response was defined as the period for which the number of IPPWs/mín remained greater than the control range. The duration of a pyloric tonic responsê was defined as the number of consecutive minules that basal pyloric prossure was > 2 mmHg. The number of contractions recorded ín all antral and duodenal side-holes were summed to provide a simple motility index.

9.2.2.5 Slatislical Analysis

Tests of significance were as described in Chapter 7

9.2.3 Results

The nasogastr¡c intubation, and the infusions of intraduodenal dextrose and intravenous atropine, were well tolerated in all but one of the volunteers. This subject (subject 7) reported transient nausea and dizziness for 10 minutes during the second dextrose infusion, and it was unclear whether this was related to the dextrose or the atropine. None complained of unexpected side effects from atropine. All volunleers developed a dry mouth (median score 7/10 after 10 minutes and 9/10 after 30 minutes), and a suslained tachycardia, with an íncrease in median heart rate from 69 (63-88) to 110 (1OO-120) after atropine (p<0.01). Mydriasis was not detected in any subject, but one complained of blurred vision 15 minutes after the third dextrose infusion had ended.

The duration of the dextrose infusions varied slightly among subjects, from 17-20 minutes (median 19 minutes), depending on the time taken for a pyloric phasic response to occur during the first dextrose infusion (see melhods). The position of the manometric assembly was well mainlained in all studies, with TMPD measurêments indicating correct location of the sleeve 107 across the pylorus for 1345 out of 1445 minutes (93%) of recording time. Because of complete loss of sleeve position in one subject, however, measurements of IPPWs and basal pyloric pressure after the third dextrose infusion were available in only 9 of the 10 subjects.

9.2.3.1 Pyloric Motor Responses to lntraduodenal Dextrose The first dextrose infusion was associated with stimulation of IPPWs and an increase in basal pyloric pressure (Figure 9.2).

(1) lsolated pyloric pressure waves The group dala for the IPPW responses to the three dextrose infusíons are illustrated in Figure 9.3. During the first infusion, the median number of IPPWs reached a maximum of 1.6 IPPWs/min,5 to 10 minutes after the dextrose reached lhe duodenum (p = 0.002). The IPPW response had a median duration of 22 minutes (7-35) after the end of the dextrose infusion, and its lotal duration was 42 minutes (30-45). The IPPW response to lhe second dextrose infusion, before atropine was administered, was not significantly different to the first infusion (0.8 vs 1.3 IPPW/min). After administration of atropine (Figures 9.3 and 9.5), there was a reduction in IPPWs (p<0.05) in I of the 1C subjects.

The lhird dextrose infusion failed lo elicit significant stimulation of IPPWs in any subject (Table 9.1), including the two volunteers in whom there was no initial response to atropine in the second infusion (subjects 6 and 7).

(2) Basal pyloric pressure During all control periods, median basal pyloric pressure was minimal (<1mmHg). Basal pyloric pressure (Figure 9.4) rose significantly within 5 minutes of the start of the first dextrose infusion (p<0.01), reaching a peak of 4.9 mmHg (3.3 - 9.8) after 1 5 - 20 minutes. The duration of this response was 35 minutes, persisting 15 minutes after the end of the dextrose infusion.

During the second dextrose infusion, the increase in basal pyloric pressure before atropine administration was not significantly different from the response to the first infusion (3.1 mmHg vs 2.3 mmHg). After atropine, there was a reduction in basal pyloric pressure in 6 of the 10 subjects (Figure 9.5), and the median change did not achieve statistical significance (p=0.2). The duration of the tonic response was 25 minules during the second dextrose infusion, compared lo 35 minutes during the first (p<0.05). 108

mmHg 40

I mlnuta

Flgure 9.2 The pyloric motor response to an intraduodenal dexlrose infusion is shown in this continuous 15 minute recording from the sleeve sensor. Within 5 minutes of the start of the dextrose infusion there is an elevalion of the basal pyloric pressure, together with the onset of phasic pressure waves isolated to the pylorus (lPPWs). 109

INFUSION 1 INFUSION 2 tNFUS|ON 3

Atrooine 2 2 2 Boíus lnf usion At rne lnfusion IPPWs tlllt No. per minute

0 0- 0 257" Dextrose v7777777777J

-10 0 10 20 30 -10 0 10 20 30 -10 0 10 20 30 TIME minutes

Flgure 9.3 lsolated pyloric pressure wave response to the first, second and third intraduodenal dextrose infusions. Group data are represented as median with interquartile range during control periods and for 30 minutes after the start of the test infusion (hatched bars). The t¡m¡ng of administralion of the intravenous atropine bolus varied slightly between subjects, from 6 to 13 minutes, and this is shown by five representative vertical arrows. 110

INFUSION 1 INFUSION 2 INFUSION 3

Atropine 10 10 10 Bolus lnfusion A ne lnfusion

Basal Pylorlc S 5 5 Pressure

mmHg 0 0 0 257" Dextrose w77777777Ã W 177777777Ã

-10 0 10 20 30 -10 o 1o 20 30 -10 0 10 20 30 TIME minutes

Flgure 9.4 Basal pyloric pressure response to the first, second and third intraduodenal dextrose infusions (medians and interquartile ranges) for the control period and for the first 30 minutes of recording after the start of the test infusions (hatched bars). 111

IPPWs BASAL PYLORIC PRESSURE

no./min mmHl * * 6 2

I I 2

I 0 0- PRE. POST. PRE. POST. ATROPINE ATROP¡NE ATROPINE ATROPINE

Figure 9.5 lndividual data showing the effect of the atropine bolus on the phasic (lPPWs) and tonic (basal pyloric pressure) pyloric motor response 1o intraduodenal dextrose. Median values are indicated by the horizontal bars. The median reduclion in IPPWs after atropine was significanl al the p<0.05 level. One subject (*) experienced transient nausea after atropine during the second dextrose infusion. 112

INFUSION 1 INFUS¡ON 3 (lPPWs/min) (lPPWs/min)

Subject Control 0-30min Control 0-30min

1 0.4 1.1 2 0 1.0 0 0 3 0 0.6 0 0.03 4 0.1 1.0 0 0.03 5 0 0.9 0 0.06 6 0 0.6 0. 2 0.06 7 0.1 2.6 0 o.2 I 0 1.0 0 0.09 9 0.3 1.4 0 o.4 10 0 2.O 0 o.o7 t Median 0 1.0 0 0.06 (lnterquartile range) (o-0.1) (o.e-1.4) (0-0) (0.03-0.15)

Table 9.1 lndividuai subject data for the isolaled pyloric pressure wave (IPPW) response before (lnfusion i) and during (lnfusion 3) atropinization. Data are presented as mean values for the 10 min control periods and for 30 min after the start of the dextrose infusions. 'p<0.01 compared to infusion 1 (0-30 min)

During the third dextrose infusion, the increase in basal pyloric pressure was delayed (5-10 minules vs 5 minutes) and significantly smaller than the increase during the first dextrose infusion (p<0.01) (Figure 9.4).

9.2.3.2 Antroduodenal Motor Changes (1) Antral motility The first intraduodenal dextrose infusion was associated with a reduction of the antral motility index in 3 of the 4 subjects in whom antral pressure waves were recorded in control periods. ln the remaining 6 subjects, and in all subjects after alropine, antral pressure waves were recorded in only 5 out of the 203 5 minute periods, and were too infrequent to allow statistical analysis (Figure 9.6). 113

|NFUSION 1 INFUSION 2 INFUSION 3

I Duodenal 10 q A Antral ton Atropin Mot lllty lnfusion lndex 5 pw/min

0 0 257" Dextrose v777777774 v7777777777Å

-10 0 10 20 30 -10 0 10 20 30 -10 0 10 20 30 TIME minutes

Flgure 9.6 Aniral and duodenal pressure waves during the first, second and third intraduodenal dextrose infusions (medlans and interquartile ranges). Atropine resulted in a reduction in antral and duodenal motility as well as in the number of propagated antropyloroduodenal pressure waves (not shown). The reduction in duodenal motility was not statistically significant. 114

(2) Duodenal motility There was an increase in duodenal pressure wavss within lhe first 5 minutes of inlraduodenal dextrose infusion (Figure 9.6), which was recorded predominantly ¡n two most distal duodenal side-holes. On 2 occasions brief episodes of phase lll-like duodenal motor activity occurred. Although there was a reduction in the number of duodenal pressure waves after the atropine bolus (p<0.01), there was no overall difference in duodenal motility during the first and second dextrose infusions, and the reduction may not have been due to alropine. The duodenal motility index during the third dextrose infusion was less than during the first infusion, but this difference was not significant.

(3) Propagated pressure waves Eighteen propagated pressure waves involving lhe antrum, pylorus or duodenum were recorded before atropine. After atropine, no pressure waves fulfilled the criteria for propagation.

9.2.4 Summary of Results The effects of intravenous atropine on the antral, pyloric and duodenal motor responses to intraduodenal infusions ol 25cr," dexlrose were assessed in 10 normal volunleers. ln each experiment, a series of 3 intraduodenal infusions ol 25!" dextrose was given. The first infusion stimulated localized phasic and tonic increases in pyloric pressure. During the second dextrose infusion, inlravenous atropine was given as a bolus (15 mcg/kg) followed by an infusion (4 mcg/kg/hr) which was continued until the end of each experiment. Before atropine was given, the pyloric motor response to the second dextrose infusion was not significantly different to the response to the first infusion, but after administration of atropine there was a rapid decrease in the rate of isolated pyloric pressure waves from 0.8 to 0.1 per minute (p<0.05). The isolated pyloric pressure wave response to the third dextrose infusion was completely blocked, and there was a much smaller maximum increase in basal pyloric prêssurê compared with the first infusion (p<0.01). This study indicates that intraduodenal dextrose reproducibly stimulales isolated pyloric pressure waves and increases basal pyloric pressure by mechanisms that involve muscarinic receptors.

9.3 A STUDY OF THE EFFECT OF DISTAL SMALL INTEST¡NAL TRIGLYCERIDE INFUSION ON PYLORIC, ANTRAL AND DUODENAL MOTILITY, GASTRIC EMPTYING AND THE INTRAGASTRIC DISTRIBUTION OF A SOLID MEAL

9.3.1 Background The presence of nutrients such as triglyceride emulsions and complex carbohydrates in the terminal ileum has been shown to retard gastric emptying and small bowel transit (Chapter 5)' This "ileal brake' may be important in the normal control of nutrient absorption and gastric emptying (Read et al 1984; Spiller et al 1984; Holgate and Read 1985; Spiller et al 1988; Welch 115

et al 1988; Jain et al 1989). Also, this mechanism may be partly responsible for the delay in mouth-caecum transit observed in some patients with steatorrhoea (Spiller et al 1987). As there have been no studies of the gastroduodenal motor patterns produced by aclivation of the ileal brake mechanism, it was our aim to do this in the present study.

9.3.2 Materlals and Methods 9.3.2.1 Subjects Studies were performed on I healthy volunteers (4 men, 4 women) with a mean age of 21 years (range 19-26).

9.3.2.2 Experimental Prolocol The morning afler an overnight fast, a manometric assembly designed specifically for these studies (Figure 9.7) was passed through an anaesthetized nostril. This assembly included a sleeve sensor for measurement of pyloric pressures, and a series of side-holes to record pressures in the antrum and duodenum (Chapter 7). The calheter extended beyond lhe assembly for 150cm, and incorporated an infusion port, a litanium weight and a silicon balloon at its most dislal end.

After intubation, the subject lay on the right side to facilitate the passage of the terminal end of the catheter into the duodenum, which was detected by monitoring of transmucosal potential difference (TMPD) via the distal infusion port. Entry into the duodenum was indicated by a change in TMPD from more negative than -20 mV (lumen with reference to subcutaneous electrode) to more positive than -15 mV. After this had occurred, transit of the manometric assembly through the small bowel was stimulated by inflation of the balloon with 10 ml of air. The catheter was allowed to pass down the small intestine until the distal end was 210 cm from the tip of the nose. The position of catheler with the tip in the terminal ileum and the sleeve sensor near lhe gastroduodenal junction was confirmed with fluoroscopy, and the location of the manometric assembly was then fine-tuned until TMPD measurements from the side-holes at either end of the sleeve indicated that the sleeve was situated accurately across the pylorus. An indwelling cannula was inserted into a left forearm vein for the sampling of blood during the experiment. This initial phase of the experiment took between 185 and 245 minutes to complete, so thal acquisition of gastric emptying and manometric data commenced between 1230 and 1430 hours.

With the subject seated in front of a gamma camera, normal saline (NS) was infused into the ileum at a rale of 1 ml/min (Figure 9.8). After 10 minutes, each subject ate a standardized solid test meal, identical to that described in Chapter 7. When approximately 25% of the meal had emptied from the stomach, the distal ileal NS infusion was changed to a triglyceride emulsion

(lntralipid 20%, Pharmacia (Australia) Pty. Ltd.; containing fractionated soy bean oil 200 g/1, fractionated egg phospholipid 12 gll, glycerol 22.5 g/l; ph 7.5, energy content 2 kcal/ml), which 116

SLEEVE

TMPD TMPD SIDE H

BALLOON

w T

ILEAL INFUSION TERMI ILEUM PORT

Flgure 9.7 Schematic representation of the recording assembly, with the sleeve positioned across the pylorus, and the infusion port in the terminal ileum, 210 crn from the nose. 117

MEAL 25!" 1009 Ground Beef Emptylng 1-1.5 mCi 99m-Tc Saline Flush t I + sallne se ne lleal lnfuslon 1.0 ml/mln

-45 0 45 90 TIME minutes

Flgure 9.8 Schematlc representatlon of the experlmental protocol. Pyloric, antral and duodenal mot¡lity, and gastric empty¡ng were monitored contlnuously throughout lhs study. 118

was infused al a rats of 1 ml/min for the next 45 minutes. After this time, the dislal ileal infusion was changed back to NS, a 20 ml "flush' of NS was given over 5 minutes, and the NS infusion was continued at 1 ml/min until the end of the study. Subjects were blinded 1o these changes in the ileal infusion. Venous blood samples were taken at intervals for future estimation of gastro-intestinal hormones. The duralion of the recordings after ingestion of the test meal ranged from 132 1o 330 minutes (median = 194 minutes).

9.3.2.3 Manometric Methods The disposition of recording points, recording techniques and methods for the measurement of TMPD were similar to those described in Chapter 7.

9.3.2.4 Data Analvsis (i) Analysis of pressure recordings Records were initially analysed in 15 minule periods from the completion of meal ingestion. For the purposes of presentation, data are displayed in the figures for 45 min before, during and after the triglycêride infusion. ln each study, the mean number of pressure wavês per minute and the mean basal pyloric pressurs (mmHg) were calculated for each 15 minute period.

(ii) Measurement of gastric emptying Emptying of the ee'Tc-labelled solid meal was monitored with a gamma camera linked to a dedicated minicomputer (Chapter 7). Data were collected continuously from the onset of food ingestion, but for the purposes of analysis, time zero was considered the time of meal completion. Various indices were derived from the gastric emptying curves for subsequent statistical analysis. These included: (1) the percentage of the total meal remaining in the total, proximal and distal stomach at -45, -30 and -15 min before, and 15,30,45,60,75 and 90 min after the commencement of the tr¡glyceride infusion; (2) the time before any food left the stomach, that is, the duration of the lag phase (LP); (3) the rate of linear emptying between the end of the lag phase and the commencement of the ileal triglyceride infusion; and (4) the time interval between the commencement of the ileal tríglyceride infusion and the cessation of omptying from the stomach, which was defined as the start of a 15 minute period during which less than 1% total gastric emptying occurred.

9.3.2.5 Statistical Analysis Data were analysed with the Wilcoxon matched pairs signed-rank test and Spearman's rank correlation coefficient, however, since reductions in antroduodenal mollity and an increase in pyloric contraction were predicted, one-tail tests of significance were used in this study. As elsewhere in this thesis, a p value of <0.05 was considered significant for all analyses. 119

9.3.3 Results 9.3.3.1 Subject Tolerance and Assembly Position Nasogastric intubation, ingestion of the lest meal and the ileal triglyceride infusion were all well tolerated. ln particular, no subject experienced nausea or vomiting at any stage. The position of the manometric assembly was satisfactory in all studies, with the sleeve correctly located across the pylorus Íor 91"/" of recording time.

9.3.3.2 Gastric Emptying Profile A representative gastric emptying profile from one subject, in which emptying ceased in the second 15 minute period after the start of the triglyceride infusion, is illustrated in Fígure 9.9. For the group, the median duration of the lag phase (LP) was 38 minutes (range 12 to 81), and the median rate of linear emptying between the end of the LP and the start of lhe ileal infusion was 0.41% per min (range 0.36 to 0.9). Median gastric emptying in the 45 minutes before the ileal triglyceride infusion was 2O.O"/" (18.8 - 24.3), and this slowed lo 12.61" (9.0 - 15.1, p<0.01) during the triglyceride infusion (Table 9.2). Because there was a latency of 15 lo 30 minutes before gastric empty¡ng slowed significantly (Fig.9.10), the greatest effect on emplying was seen in the 45 minutes after the end of the triglyceride infusion, when O.7% (-1.5 - 2.8) of the meal emptied (p<0.01 cf both the 45 min before and during the ileal triglyceride infusion) (Table 9.2). ln only one subject did gastric emptying recommence after the lipid infusion. This delay in gastric emptying was associated with redistribution of part of the solid meal to the proximal gastric region-of-interest (ROl) in 7 ol the 8 subjects (Figures 9.11 and 9.12), there being an increase in counts in the proximal stomach from 30 to 45min after lhe slart of the ileal lipid infusion (p<0.05), with a corresponding decrease in distal stomach counts.

BEFORE TG TG AFTER TG TIME (min) -45 to 0 0to45 45 to 90

+ Total Gastric 2 0.0 12.d -0.7 Emptying (%) 18.1-24.3 9.0-1s.1 -1.s-2.8

Table 9.2 Group data for gastric empty¡ng before, during and after terminal ileal lriglyceride (TG) infusion. Median *p.0.05, values and interquartile ranges. +pcO.01 cf before triglyceride. 120

LIPID + TOTAL 100 O PROXIMAL + DISTAL s z o zt- 50 l-l¡J UJ E

0 -100 0 100 200 TIME (min)

Flgure 9.9 Gastric emptying curve from one subject demonstraling the effect of terminal ileal triglyceride infusion on nt'Tc-labelled gaslric emplying and the intragastric distribution of a lOOg chicken liver/ground beef burger. Gastric emptying proceeded in a linear fashion after an initial lag phase. Emptying from the total gastric ROI ceased approximately 20 min after the start of the ileal lriglyceride infusion. Following the cessation of emptying, counts in the distal ROI continued to fall gradually, as part of the test meal moved retrogradely into the proximal slomach. During this time, there was a corresponding increase in proximal gastric counts. 121

16.6

12 LIPIO

TOTAL I GASTRIC EMPTYING I $t 4 I

0 t É+

-45 0 45 90

TIM E (m in)

Flgure 9.10 Group data for the p€rcent of the solid meal emptied each 15 min for 45 min before, during and after terminal ileal triglyceride infusion. 122

il .Ðr

Figure 9.11 movement of part ol the solid meal scintigraphic images from one subject demonstraling the retrograde gastric region is divided into proximal and distal associated with terminal ileal triglyceride infusion. Tie regions of interest. 123

Proxlmal Stomach

100 I LIPID I t I 80 ,a I I I tttaaa--- I b tt---z-t. 60 I ta Y' I a I RETENTION t I 40 I I c 20 d

h 0

-45 045 90

TIme minutes

D¡stal Stomach

,l,i rq 100

i h 80

d 60 g % f RETENTION c 40 a e b 20

{5 04s 90 Time minutes

Figure 9.12 Changes in scintigraphic counts in the proximal and distal gastric regions of interest for each individual subject (a to h). Median total gastric emptying is indicated by the hatched line for comparison. The reduction in counts from both regions which accompanies gaslric emptying slows in response to ileal triglyceride infusion. ln 7 subjects there is a slight increase in counts in the proximal slomach during and after the lriglyceride infusion as some of the solid meal moves retrogradely from the distal slomach. k 124

9.3.3.3 Antropyloroduodenal Motor Response

The infusion of triglyceride into the terminal ileum was associated with the stimulation of IPPWs

and a decrease in antral and duodenal pressure waves (Figures 9.13 and 9.14). Numerical data is presented in Table 9.3.

(i) lsolated pyloric pressure waves (lPPWs) lleal triglyceride infusion was associated with an increase in the rate of IPPWs in 7 of the I subjects (p=0.01). Median values are shown in Table 9.3. This stimulation of IPPWs had a latency of 15 minutes, after which the maximum rate of 1.5 lPPWmin was recorded (Figure 9.14). Although there was a decrease in the rate of IPPWs after the end of the triglyceride infusion, this decrease was nol significanl, and the rate of IPPWs in lhe 45 minutes after triglyceride remained elevated compared to the 45 minutes before triglyceride (pcO.O5).

(ii) Basal pyloric pressure

Median basal pyloric pressure remained close to zero throughout the study (Table 9.3). There was a small rise in basal pyloric prossure after the ileal triglyceride infusion in 6 of the 8 subjects, but the change was not statistically significant (p.0.1).

(iii) Antral motility

.,| rr,l ln 7 subjects distal antral pressure waves decreased significantly within 30 minutes after the 'lt, I start of the ileal triglyceride infusion. After this time few antral pressure waves were recorded. Median data for anlral, duodenal and propagated motility are shown in Table g.3.

iv) Duodenal motility

The duodenal motility index decreased significantly (p<0.05) in all subjects 1Smin after the start of triglyceride infusion (Figure 9.14). This decrease was sustained until the end of each study.

(v) Propagated pressure waves

The number of pressure wav€,s, recorded in antral, pyloric or duodenal side-holes, which were propagated aborally decreased 15 minutes after the start of the triglyceride infusion (pcO.05). This was not simply due to the overall reduction in antral and duodenal activity after the infusion, since the proportion of this activity which satisfied the criteria for propagation also decreased (p<0.01) within the f¡rst 15 minutes after the start of the lipid infusion. t 9.3.3.4 Correlation Between I Gastric Emptying and Motility Comparison of the group data each minute ; in 15 interval before, during and after the ileal triglyceride infusion demonstrated a strong correlation between the median rate of total gastric emptying and the median number of antral pressure waves (r=o.92, p<0.001), duodenal pressure

I 125

BEFOR E DURING TRIGLYCERIDE TRIGLYCERIDE

mmHg 40 I DISTAL ANTRAL SIDE.HOLE o 40 SLEEVE o

tl 40 Iil PROXIMAL . l.l ,1 DUODENAL I SIDE.HOLE o ^ ++¿,-v ¿'- L/^\ a1

3O sec¡

Flgure 9.13 Manometric traclng showing an example of the effect of terminal ileal triglyceride infusion on the pattern of coordinated pyloric motility during emptying of the solid meal.

r 126

Lipld 2

Dlstal Antral Slde-ho le pressur9 waves/min

-45 0 45 90

Sleeve IPPWs/min

-45 0 45 90

Prox lm a I Duodenal Slde-hole pressure waves/min 0-

-45 0 45 90 TIME minutes

FIgure 9.14 Group data for distal antral (antral TMPD side-hole), pyloric and proximal duodenal (duodenal TMpD side- hole) for 45 minutes before, during and after ileal triglyceride (lipid) infusion. Median values and interquartile range. 127

BEFORE LIPID LIPID AFTER LIPID TIME (min) -45 lo 0 0to45 45 to 90

* ANTRUM 0.58 0.30 o. oo+ t t IPPWs o.73 1.33 1.O2

PYLORICTONE 0.0 -o.7 0.8 * ¡ DIJODENIJM 2.93 1.55 o.37'

* PFIOPAGATED 1.O2 o.23 o. 04+ WAVES

Table 9.3 Group data for motility (number of pressure waves/min) before, during and after terminal ileal triglyceride (lipid) infusion. Median values and interquartile range. *pcO.OS, +p<0.01 cf belore lipid.

waves (r=0.80, p<0.01), and propagated pressure waves (r=0.88, p<0.0025). There was an inverse relationship between gastric emptying rate and IPPWs (r=-0.60, p=0.05). Although there was little increase in basal pyloric pressure (0.8 mmHg), there was a strong inverse relationship between basal pyloric pressure and gastric emptying (r=-0.90, p<0.0025).

9.3.4 Summary of Results ln this study, the effects of a terminal ileal triglyceride inf usion on antropyloroduodenal pressures and gastric emptying were recorded in 8 healthy volunteers after ingestion of a solid meal. Gastric emptying slowed markedly 15-30 minutes after the start of the lipid infusion (p=O.Ot ) and, in lhe majority of subjects, lhere was retrograde movement of the solid meal from the distal to the proximal stomach. During the lipid infusion, there was a decrease in antral (p=0.01), duodenal (pco.os) and propagated antropyloroduodenal pressure waves (p<0.05), and an increase in isolated pyloric pressure waves (p<0.05). The rate of gaslric empty¡ng correlated with antral pressure waves (r=0.92, p<0.001 ), duodenal pressure waves (r=0.80, p<0.01), and propagated pressure waves (r=0.88, p,0.0025), and inversely with the number of isolated pyloric pressure wavos (r=0.60, p=0.05). The changes in antral, pyloric and duodenal motilily, 128

and the intragastr¡c redistribution of a solid meal associated with ileal lipid infusion, are likely to contribute to lhs delay in gastric emptying caused by this stimulus.

9.4 DISCUSSION The major new findings of these two studios on nutrient induced changes in pyloric motility, are that i) lhe pyloric motor response to intraduodenal dextrose infusion in humans is atropine- sensitive, and ii) stimulation of distal small bowel receptors in humans induces a coordinated pattsrn of pyloric motility similar to that induced by proximal small bowel stimuli which also retard gastric emptying. ln addition, it was shown in the first study that IPPWs and pyloric tone are reproducibly stimulated by dextrose infusions which deliver calories into the duodenum at a rate similar to the known calorie-limited rate of emptying of glucose from the stomach (Brener et al 1983). This confirms previous findings (Heddle et al 1988c) and lends weight to the concept that the pylorus is important in the physiological control of gastric emptying. Finally, it has also been demonstrated for the first time that the delay in gastric emptying after ileal triglyceride infusion is often associated with retrograde movement of food within the stomach.

Recent studies have demonstrated that the intensity and duration of the pyloric motor response to intraduodenal dextrose infusions is dependent on the concentration of the dextrose infusate and, in addition, that this response is reproducible after repeated stimulation with consecutive infusions (Heddle et al 1988c). The mechanism by which the dextrose infusate interacts with small bowel receptors remains unclear. Previous experiments indicate that, while the osmolarity of the solution is of importance, the calorie contsnt per se appears to provide a substanlial additional stimulus (Heddle el al 1988c; Chapter 5). ln the study reported in this chapter, atropine was tested against the pyloric motor response to a concentration of dextrose which had been previously shown 1o be most effective in causing this sÌimulated manometric pattern of IPPWs and pyloric tone. The abolition of IPPWs and the reduction in pyloric tone observed in the present study can therefore be attributed to the administration of atropine.

What is the significance of the muscarinic dependency of these changes in pyloric motility? This finding narrows the field of possible mediators of the pyloric motor response to small bowel receptor stimulation by dextrose. Atropine will block the effects of both acetylcholine released from post-ganglionic nourones, and hormones acting via muscarinic receptors. ln dogs, the pyloric motor response to intraduodenal acid is mediated by ascending, intramural cholinergic excitalory neurones (Allescher et al 1988), and our findings are consislent with such a pathway being present in humans as well. A similar pathway apparently exists in the cal, although it may involve opiate rather than cholinergic receptors (Reynolds et al 1984 & 1985). Other neural pathways may also be important, however. For example, the vagus may provide a cholinergic 129

excitatory input, since this has already been demonstrated after electrical stimulation of the vagus in dogs (Allescher et al 1987; Mir et al 1979; Chapter 5).

The findings are also consistent with mediation of the dextrose-induced response by hormonal feedback to the pylorus (Chapter 5). Various gastrointestinal hormones are released after ingestion of carbohydrate-containing mixed meals. Glucose is known to stimulate the release of gastric inhibitory polypeptide (GlP) and insulin (Brown et al 1975), but it is difficult to be sure if glucose independently stimulates the release of other hormones, because plasma levels of hormones have usually been studied after ingestion of mixed meals, containing other constiluents in addition to carbohydrate. Gastric emptying is slowed by apparently physiological concentrations of a number of gastrointestinal hormones (Chapter 5), including cholecystokinin (CCK) (Debas et al 1975), secretin (Kleibeuker et al 1988b), peptide YY (PYY) (Pappas et al 19g6), gastrointestinal inhibitory polypeplide (Brown et al 1975) and glucagon (Phaosawasdi and Fisher 1g82). The effect of these hormones on pyloric motil¡ty, however, has not been analysed extensively. Cholecystokinin has been most studied (Section 5.5.1), and at physiological levels increases basal pyloric pressure and stimulates phasic pyloric contractions in dogs and rats (Scheurer et al 1983; Telford et al 1979). lts effect on canine gastric emptying is reduced by pyloroplasty (Yamagishi and Debas 1978). The excitatory effects of cholecystokinin on gastric smooth muscle in the guinea-pig and dog are abolished with atropine, indicating that cholecystokinin can act via acelylcholine release (Gerner and Haffner 1977), but such an action has not been established for the pylorus. Although the evidence is conflicting, cholecystokinin concentrations have been shown to increase significantly after oral glucose (Liddle et al 1985; Wiley et al 1988), so cholecystokinin is a candidate hormone for mediating the effect of dextrose on pyloric motility. Cholecystokinin is perhaps more likely to be a fat-induced inhibitor of gastric emptying (Kleibeuker et al 1988a), also via changes in pyloric motility (Tougas et al 1987; Heddle et al 1988b). Secretin also induces pyloric contraction in animals (Lipshutz and Cohen 1972; Ruckebusch and Malbert 1986) and possibly in humans (Fisher et al 1973), but iÎ is not yet clear whether the release of secretin is stimulated by glucose. The importance of other gut hormones, particularly GIP and PYY, in the dextrose-induced pyloric motor response, requires further evaluation.

The reason for the differential effect of atropine on IPPWs and basal Pyloric pressure observed in this study is unclear. Pyloric tone may still be acêtylcholine dependent, but relatively atropine- resistant. The bolus dose of atropine chosen in this study was similar to doses used previously to investigate gastrointestinal motility and gastric emptying (Valenzuela el al 1976; Jaup and Dotevall 1981; Wiley et al 1988), and it was believed that the subjects would find it tolerable. The continuous atropine infusion was designed, using data on the elimination half-time of atropine (Adams et al 1982; Dent unpublished), with the aim of maintaining the maximum plasma levels 130

initially achieved by the bolus. lt is possible, nevertheless, that lhe partial effect of atropine on stimulated basal pyloric pressure was simply a reflection of incomplele atropinization. An alternalive explanation is that muscarinic receptors may be responsible for only part of the tonic rssponse, and lhat lhe pathways mediating IPPWs and increased basal pyloric pressure may differ. Thus, while muscarinic receptors appear to be involved at some level in the stimulation of pyloric tone after dextrose, acetylcholine may not be lhe final transmitter.

There have been few previous studies in humans of the pathways mediating the pyloric motor responses to duodenal receptor stimulat¡on. Enkephalin, as well as olher neuropeptides such as substance P, have been demonstrated within nerve fibres of the human pylorus (Wattchow et al lggg), but lhe physiological role of these substances at Present remains unknown. Bovell et al (1gg7), using a sleeve sensor, found no effect of naloxone on pyloric contraction induced by intraduodenal infusions of fat. Dooley et al (1985) concluded that the pyloric pressure resPonse to duodenal acidification was prevented by naloxone, but the posilion of the sleeve sensor in that study was not adêquately monitored, and their recordings of stimulated pyloric motility are at variance with those in the present study. Valenzuela et al (1976) recorded gastroduodenal motility with side-holes, and suggested that lhe pyloric motor rssponss to intraduodenal acid infusion was atropine sensitive. Although lhese findings are eonsistent with the present study, there remains some doubt about their validity, since sampling times were short and the manomelric lechnique was unreliable (Chapter 2).

Can the effect of atropine on the IPPW response give insight into the imporlance of the pylorus in the control of gastric emptying? Pyloric motility is only one of a number of motor mechanisms which act in concert to regulate gastric emptying (Meyer 1987). ln particular, alterations in gastric fundal pressure, suppression of antral pumping and reductions in duodenal motility are all likely to be important contributory factors. lf the effect of atropins on gastric emptying was dominated by the suppression of pyloric contraction, one would expect atropine to accelerate gastric emptying. Atropine, however, delays gastric emptying of non-liquid meals (Subissi et al 1g85), but this may be due to reductions in antral motility, proximal stomach lone and inlragastric pressure. The pylorus may be relatively more important in the control of nutrient liquid emptying than for the emptying of non-nutrient liquids (Chapter 6) and, if this were so, atropine should increase the rate of emptying ol 251" dextrose. Allhough to the author's knowledge this has not been studied, it is interesting that truncal vagotomy (without pyloroplasty) may lead to a more rapid emptying of hypertonic glucose meals in humans (Hall and Read 1970), whereas the emptying of hypolonic NaCl is not affected. lf vagotomy had an effect purely on the proximal stomach (receptive relaxation) or antrum, one would not expect this differential effect. The authors of that study concluded that duodenal osmoreceptors acted by being connected to lhe vagus. 131

ln the second set of experiments reported in this chapter, ileal lriglyceride infusion resulted in marked slowing of gastric emptying in all subjecls, confirming the findings of previous studies (Holgate and Read 1985; Welch et al 1988). The changes in antropyloroduodenal motilily associated with this potent delay in gastric emptying were (1) stimulation of the pylorus, and (2) suppression of antral and duodenal motility. The isolated pyloric pressure waves induced by ileal triglyceride infusion were similar to the characteristic motor Pattern recorded in the first study after stimulation of duodenal nutrient rocaptors by dextrose. This pattern has been well recognized by other investigators, and has also been shown to be associated with delayed gastric emptying after intraduodenal triglyceride infusions (Chapters 4, 5 and 6). Smaller elevations of basal pyloric pressure were recorded after ileal triglyceride infusion when compared to previous studies with intraduodenal triglyceride or dextrose (Heddle et al 1988b&c). This apparent difference in tonic response, however, may reflect a type 2 error due to the relatively small number of subjects in the present study. ln view of lhe otherwise close similarity of lhe motor responses to ileal triglyceride and a variety of duodenal nutrients, it is possible that these retardant stimuli may act via a final common pathway'

There was a lalency of 15-30 minutes between the slart of the ileal lipid infusion and the onset of the changes in both motility and gastric emptying. This delay, which was observed in all subjects, could be due to the time taken for the lipid infusion to recruit the required number of mucosal fat receptors close to lhe infusion site (Lin et al 1989a&b), for the lipid to pass along fhe ileum before it reached the receptors, or possibly for mediation of the effect after ileal sensing. The existence of a latency therefore does not necessarily favour the release of gastrointestinal hormones, rather than activation of neural pathways, as the mechanism mosl likely to mediate the ileal brake. The reductions in antral and duodenal motility and the slowing of gastric emptying were also sustained well after the end of the ileal fat infusion' Possibly, despite the saline 'washout" after the fat infusion, this could have resulted from continued stimulation of ileal mucosal receptors by triglyceride within the gut lumen. Alternatively, the findings could be explained by prolonged elevation of neurotransmilters or gastrointestinal hormones, especially peptide YY, acting at a local level or possibly via the central nervous system (Allen et al 1984).

The second major finding of this sludy was the retrograde movement of the solid meal as gastric emptying slowed. This redistribution of food from the dislal to the proximal stomach was small but clearly apparent in the majority of subjects. Alterations in gastric fundal tone, thought to be important in the control of gastric emptying (Aspiroz and Malagelada 1985; Meyer 1987), could have contributed to this movement, though an increase in antral tone is also possible. More localized changes of tonic gastric contraction might also be important, however. Previous radionuclide studies have demonstrated a mid-gastric contraction band which becomes evident as solid food moves distally during normal gastric emptying (Moore et al 1986; Collins et al 1988). 132

A similar band was seen in some subjects in the present study, and was associated with the retrograde movement of the solid meal as emplying slowed. Such a band could participate in the control of gastric emptying, in concert with changes in regional gastric tone, by retarding the movement of food from the proximal to the distal stomach.

These effects on gastric emptying and gastroduodenal motility were achieved using an infusion which delivered a calorie load to the ileum at the lower limit of the normal rate of entry of lriglyceride into the upper small intestine (Hunt and Stubbs 1975). Studies on ulcerative colitis patienls with ileostomies indicate that appreciable amounts of fat can pass through the small intesline (Neal et al 1984). There is also indirect evidence from these patients that stimulation of ileal receptors normally modulates upper gastrointestinal function, since mouth - stoma transit times were slower in those with an intact ileum than in those who had undergone ileal resection. The potency of the effect of ileal triglyceride infusion at the levels used in the present study, together with the above observations, suggests that the ileal brake may operate under physiological conditions. Although ileal protein or glucose infusions do not app€ar to delay gastric emptying (Welch et al 1988), activation of the ileal brake has been demonstrated after infusion of complex carbohydrates into the ileum at levels considered to be physiological (Jain et al 1989). Regardless of the stimulus, however, the ileal brake is likely to be most complelely activated in states of malabsorption.

Resulls from these sludies indicate that pyloric motility is influenced by stimuli distant from the gastroduodenal junction. The changes in motility associated with the delay in gastric emptying by cold stress, which resemble those observed with ileal triglyceride, are reported in the following chapter. There has been no assessment of the motor effects of other stimuli acling at sites distant from the pylorus, such as rectal distension, a stimulus which has also be'en reported to modulate upper gastrointestinal function (Youle et al 1984; Kellow et al 1987). However, the similarities among the gastroduodenal motor responses to ileal fat, strsss and intraduodenal nutrients suggest that many of the feedback mechanisms which influence the rale of gastric emptying may do so via the same pattorn of alteration of gastroduodenal motility, irrespective of the nature of the stimulus.

9.5 CONCLUS¡ON Stimulation of distal small intestinal nutrient receptors by triglyceride induces changes in the Pattern of pyloric motility which are similar to those due lo feedback from stimulation of upper small intestinal receptors by fat. This conformity suggests thal the mechanisms conlrolling the pylorus may act via a final common palhway, although none of the individual neurochemical mediators have previously been identified in humans. The finding that a muscarinic pathway is involved in upper small intestinal feedback regulation of pyloric motility indicates that atropine 133

should be tested againsl other slimuli of localized pyloric motility. Furthermore, in the light of this finding, further investigations of the hormonal control of the pylorus can be directed more appropriately. 134

CHAPTER 10

A Study of Pylorlc Motility During the Delay in Gastrlc Emptying Induced by Cold Stress

10.1 INTRODUCTION Gastrointestinal motility and gastric emptying are modified by a variety of centrally acting stressors, such as labyrinthine and psychological stimulation and cold pain (Thompson et al 1982 & 1983; Stanghellini 1983 & 1984; see also Chapter 5). Psychological stress inhibits fasting human jejunal motor activity (McRae et al 1982), but its effects on gastric emptying and small bowel transit seem variable and subject to tolerance (Cann et al 1983; O'Brien et al 1987). The cold pressor test, however, reliably delays gastric emptying (Thompson et al 1982). Other physiological effects of cold stress include altered gastric and pancreatic secretory function (Thompson et al 1982), a rise in systolic and diastolic blood pressure and initial increases in skin conductance levels and respiration (Lovallo 1985). While these responses depend on inlact peripheral innervation, they are largely mediated by the central nervous system (Lovallo 1985). The effect of cold stress on gastric emplying has been reported lo be associated with inhibition of antral motility without any significant effect on duodenal motility (Stanghellini et al 1983), but there is no information about the effect of this stimulus on the pylorus. ln the sludy reported in this chapter, the effects of cold stress on pyloric, antral and duodenal motility have been assessed, and correlated with lhe changes in gastric emptying induced by this slimulus.

1O.2 MATERIALS AND METHODS 10.2.1 Sublects Studies were performed on 7 healthy volunteers (3 males,4 females), aged 18 to 22 years. 135

MEAL 25o/o 1009 Ground Beef Emptylng 1-1.5 mCi 99m-Tc I I COLD STRESS -20 o20 40 60 TIME minutes

Flgure 10.1 Schematic representat¡on of the experimental protocol. Pyloric, antral and duodenal pressures' as well as gastric emptying, were monitored continuously throughout the study.

10.2.2 Experlmental Protocol After intubation and positioning of the sleeve sensor across the pylorus, the subject was seated in front of a gamma camera and given a standardized labelled solid meal (Chapter 7), which was ingested within 5 minutes. When approximately 25ø," of the test meal had emptied from the stomach, each volunteer was stressed by repeated immersions of the left hand into iced water (4oC) for 60 seconds, with an interval of 15 seconds between immersions when the volunteer held the hand in room air (Figure 10.1). Cold stress was continued for a maximum of 20 minutes, or for a shorter period if it proved intolerable for the subject. Measurements of motility and gastric emptying were conlinued for at least 40 minutes after cessation of the immersions. Prior lo each study, volunteers completed two quest¡onnaires to assess their susceptibility to stress: the State Trait Anxiety lnventory Self Evaluation Questionnaire (Spielberger 1983) and the 136

Maudsley Personality lnventory (Eysenck 1980). Al 20 minute intervals before and after the cold stress, and every 5 minutes during cold stress, pulse and blood pressure were recorded, and volunteers were asked to indicate separately their subjective experience of strsss, pain and hand coldness on a visual analogue scale from 0 to 10.

10.2.3 Manometric Methods The recording techniques and fhe manometric assembly were identical to lhose described in Chapter 7.

10.2.4 Data and Stat¡st¡cal Analysis Records were analysed in 5 minute periods Íor 20 minutes before, 20 minutes during and 40 minutes after the completion of cold stress. The median number of pyloric, antral and duodenal ptessure waves for each 5 minutes period during and after stress were compared to the 5 min period immediately before stress, using the Wilcoxon matched pairs signed-rank test. The number of pressure waves in each complete 20 minutes period before, during and after stross were also compared using this test. Differences in cardiovascular parameters were also assessed with the Wilcoxon test. The relationship between median values for gastric emptying and the median number of pressure waves for all 20 minules periods was assessed with the Spearman rank correlation coefficient. The mean and standard deviation of scores from the personalify questionnaires were compared to those from an appropriate control population (American college students), using Student's t test. A p value of <0.05 was considered significant for all analyses.

1 0.3 RESULTS 10.3.1 Sublect Tolerance of Cold Stress Scores from rhe Srate Trait Anxiety lnventory were not significantly different from the normal population control, with a State anxiety score ol 39.7 r/- 8.9 vs 37'7 r/- 11'0, and a Trait anxiety score of O9.7 +t- 9.2 vs 89.5 +/-9.7 (Spielberger 1983). Neuroticism and extroversion scores from lhe Maudsley Personality lnventory were slightly, but not significantly, higher than fhe population control (26.0+l- 11.3 vs 2O.9 +l-1O.7, 34.7+l-5.3 vs 28.5+/- 8.3 respectively) (Eysenck 1g8O). During the cold pressor têst, each subject indicated significant pain, stress and hand coldness on the visual analogue scale (Table 10.1). Five subjects completed 20 minutes of strsss. lmmersions were disconlinued due lo discomfort alter 1O minutes in one subject, and after 5 minutes in another who became hypotensive (systolic blood pressure 80 mmHg) and bradycardic (56/min). One other subject developed bradycardia without hypotension. Consistent with the previously reported response to the cold pressor test (Lovallo 1985), lhere was a median rise (p<0.01) in systolic and diastolic blood pressure within 5 minutes, but no change in pulse rate (Table 10.1). 137

BEFORE STRESS DURING STRESS AFTER STRESS -20 5 5 10 15 20 +5 +20 lminì lminì lminl BP (mmHg) Systolic 116 110 146' 1M' 140ú 1/15 124 126 Diastolic 74 85 got 90+ 95+ 96 80 u

Pulse/min 72 72 84 78 7s 70 70 67

g* ooLD 1 2 S+ 7+ 5+ 5 PAIN 2 2 6t g* 5+ 4+ 6 gf g* STRESS 0 1 7+ 6+ 0

Table 10.1 Effect of cold stress on blood pressure (BP), pulse and subjective parameters measured on the linear analogue scale (maximum score '10). * Median values. p<0.01 , + p<0.05 cf before stress.

10.3.2 Gastrlc Emptying The median rate of gastric emptying slowed (p<0.01) from 17"/. in the 20 minutes before stress lo 2/" in the 20 minutes during stress (Figures 10.2 and 10.3a). This significant decrease was sustained for the first (5% gastric emptying, p<0.01) but not the second (12/", p

The delay in lotal gastric emptying during stress was accompanied by retrograde movement of the meal from the distal to the proximal stomach in 4 subjects (Figures 1O.2, 10.3 and 10.4)' After stress, the rate of gastric emptying increased, so that from 20-40 minutes after the end of the immersions, lhe rate was no longer significantly different to before stress (Figure 10.3a). Median scintigraphic counls in the distal stomach did nol decrease after stress despite the increase in the rate of gastric emptying, due to increased filling from the proximal stomach. 138

120 Stress E Total o Proximal 100 o Distal

80 Retentlon 60 ofto

0 0 100 200 TIME minutes

Figure 10.2 lndividual example of the change in gastric emptying and scintigraphic counts in lhe proximal and distal gastr¡c regions of interest (ROl) with cold stress. Retrograde movement of the solid meal associated with ðtress is indicated by a decrease in distal ROI counts and a corresponding increase in proximal ROI counts. 139

A 30 STRESS

20 Gastric Emptylng % * 10

ffiffi ffi -20-0 0-20 20-40 40-60 min min min min

B 20 trl Proximal ROI E Distal ROI

10 Reductio n * ln -T Sclntlg rap h ic Counts #= olto 0 F

-10 -20-0 0-20 20-40 40-60 min min min min

TIME minutes

Figure 10.3 Effect of cold stress on gastric emptying (A) and the emptying (reduction in scintigraphic counts) from the proximal and distal gastric regions (B). The negative 25th percentile for the proximal gastric ROI during and after slress is due to an increase in counts in this region in 4 subiects, caused by retrograde movement of the meal. Values are medians with interquartile ranges. * pcO.O1 compared to the 20 min before stress. 140

movement of part of the solid 5l?,::"",tiiigraphic gasrric images from one subjecr showing rhe rerrograde interest, associaled with cold stress' meal, from the distal to t" proximal gastric region of 141

10.3.3 Pyloric Motility The application of cold stress was associated with a significant modification of the fed pattern of pyloric motility (Figure 10.5). There was a significant (p<0.05) increase in the median number of isolated pyloric pressure waves (lPPWs) during strsss, compared to the 20 minutes before stress (1.0 vs 0.3 IPPW/min respectively). There was a significant increase in IPPWs compared to before slress during the first (0.7 lPPWmin, p<0.05), but not in the second (0.6 IPPW/min, p<0.1) 20 minute period after stress (Table 10.2). The time course of these changes is shown in Figure 10.6. There was no significant change ín basal pyloric pressure.

10.3.4 Antral and Duodenal Motility 10.3.4.1 Antral Motility A significant and sustained decrease in the median number of distal antral pressure waves compared to before stress (1.S/min) was observed in the 20 minutes during stress (0.3/min,

p<0.01), and the first (1.O/min, p<0.05) but not significantly in the second (0.9 pw/min, P<0.1) 20 minute period after cold stress (Table 10.2). This decrease occurred within 5 minutes of starting cold stress (Figure 10.6). Although pressure waves were observed less frequently at the antral sideholes 2 and 4 cm proximal to the sleeve, there were also significant reductions (p<0.05) in lhe median number of pressure waves during slress compared to before stress at these sites (0.0/min from 0.25lmin, and 0.O/min lrom O.2/min respectively). The number of aborally propagated waves was reduced from a median of 1.8/min before stress, to 0.5 during (p<0.01) and 0.8 (p<0.05) and 1.3/min (NS) in the two 20 minute periods after stress (Table 10.2).

10.3.4.2 Duodenal Motility The number of pressure waves recorded each 5 minutes by the two proximal duodenal side-holes (D1, D2) was significantly reduced (p<0.01) during cold stress, compared to the 5 minutes

immediately before stress (Figures 10.5 and 10.7). ln two subjects, short episodes (3 min and 1 min 55 sec) of pressure waves al a rale of 11/min were observed 3 cm and 9 cm distal to the sleeve (D2, D3) immediately after the start of cold stress (Figure 10.8), in the absence of any associated phase lll-like activity in the antrum. Because of this, the reduction in the median number of pressure waves at these sites in the 20 minules period during stress was not significant compared to before stress (0.6 vs 0.9/min and 0.9 vs 1.7/min respectively, p<0.1 for both). The number of pressure waves recorded at the distal duodenal side-hole (D3) was not significantly different during, compared to before stress (1.1 vs 1.O/min). 142

mmHg '"1 A1

'"1 ^2

A3 i"1 a a a SLEEVE

D1 '"1 "J il '"1 '\ D2

BEFORE STRESS DURING STRESS

3O..c

Figure 10.5 stress. After stress, there is Example of the changes in antral, pyloric and duodenal motil¡ty induced by cold and duodenal pressure some artefact seen in all channels due to straining. There is suppression of antral pressure waves (asterisks)' waves; the only pressure waves shown during stress are isolated pyloric

I 143

BEFORE DURING AFTER STRESS STRESS STRESS -20 to 0 min 0 to 20 min 20 to 40 min 40 to 60 min

Dlstal antrum 1.5 (1.1-2.1) 0.3+ (0.1-0.6) 1.0 (0.s-1.4) 0.e (0.4-1.4) (Waves/min) Pylorus IPPWs/min 0.3 (0.1-0.6) 1.0 (o.2-1.21 o.7 (0.5-1.6) 0.6 (0.3-0.9) Tone (mmHg) -1.7 .1.6 -2.1 -2.O Duodenum (Waves/min) D1 0.e (0.3-1.3) 0.1' (o-0.1) 0.2 (0.1-0.5) 0.3 (0.3-1.3) t D2 r.7 (0.e-1 .e) 0.5 (0-0.6) 0.8+ (0.4-1.0) 0.e (o.7 -2.61 D3 1.0 (0.5-1.7) 0.8 (0.1-1.1) 1.1 (0.1-1.1) 1.1 (0.7-1.5)

'L Propagated waves Iti iþ (Waves/min) 1.8 (0.9-2.0 0.5+ (0.3-0.6) 0.8+ (0.5-1.0) 1.3 (1.0-1.7)

rj Table 10.2 Pyloric, antral and duodenal motility in the 20 min before, 20 min during and 40 min after cold stress. Median values and interquartile ranges.

+ pcO.01 , 'p<0.05 cf before stress

r 144

Cold Stress 3 ,i Distal Antra 1 2 Pressu re * Waves

1 No./min t la +

-20 20 40 60 * 2

lsolated * * Py lo ric P ressu re Waves No./min

.ti q -20 0 20 40 60 lt, 4

Pyloric Tone 0 mmHg

-20 20 40 60 TIME minutes

Flgure 10.6 T Group data for distal antral and pyloric motility. Medians and interquartile ranges. + p<0.01 , ' p<0.05 compared to the 5 min before stress.

I 145

Cold Stress 3 TÍ/Í//////ú.

D'! 2 Pressure Waves/min

1 *¡ * * ** * 0

-20 0 20 40 60 t I I I I 4.2 t 3 I t I t I t i t t D22 I t Pressure t Waves/min t I 1 r* *

qf -20 0 20 40 60 \i 3 6.4

D3 2 Pressure Waves/min

-20 0 20 40 60 TIME minutes

Flgure 10.7 I Group data showing the effect of cold stress on duodenal prsssure waves, for the duodenal side-holes located in the most proximal cap (D1), and 3 cm (D2) and 9 cm (D3) distally. Median values and interquartile ranges. * p<0.05 compared to the 5 min before stress. r 146

COLD STRESS mmHg.J DISTAL ANTRUM

'.{ D1

.J D2

'.1 D3

30.¡c

Figure 10.8 lndividual tracing showing an burst of duodenal phase lll-like pressure waves shortly alter the start of cold stress. Note that this phenomenon is most evident in the two more distal duodenal side'holes (D2 and D3).

I f

T I lr

r 147

10.3.5 Relatlonship Between Motility and Gastrlc Emptylng The amount of emptying from the total stomach in the 20 minutes before, 20 minutes during and the 40 minutes after stress correlated directly with the median number of distal antral pressure waves (r=0.66, p<0.0001; Figure 10.9a) and propagated pressure waves involving the antrum, pylorus and duodenum (r=0.63, p<0.0004; Figure 10.9b). Excluding the episodes of brief duodenal phase lll-like activity immediately after the onset of slress in 2 subjects, there was also a correlalion betwsen gastríc emptying and duodenal pressure waves (r=0.55, p<0.0025). There was no significant correlation between gastric emptying and IPPWs (r=0.20, p<0.1).

r= 0.66 p<0.0001 a 30 a

Gast ric 20 a Emptying a a a a fo 10 a o ; ta I ¡o :' a a o 0 -]a ,

0 0.5 1.0 1 .5 2.O 2.5 Antral Motlllty pw/min

Flgure 10.9a Relationship between the number of antral pressure waves recorded in the distal antral side-hole, and the rate of gastric emptying in the 20 min periods before, during and after cold stress 148

r=0.63 p<0.0004 3 a

Gastric 2 a Emptying o a o a I 1 I t . a l .' a ¡ a a 0 - -¡ -Qr

0 0.5 1 .0 1.5 2.O 2.5 Propagated Pressure Waves No./minute

Figure f0.9b Relationship between the number of aborally propagated waves involving the antrum, pylorus and duodenum, and the rate of gastric emptying in the 20 min before, 20 min during and 40 min after cold stress.

10.4 SUMMARY OF STUDY The changes in gastroduodenal motility respons¡ble for the delay in gastric emptying produced by cold stress were assessed in this study. Antropyloroduodenal pressures were recorded in seven healthy volunteers, concurrent with scintigraphic measurement of gastric emptying. When approximately 25% of the solid meal had emptied from the stomach, a standard cold pressor test was applied. Cold stress was associated with slowed emptying from the total stomach and lhe proximal stomach. Retrograde movement of the solid meal from the distal to the proximal stomach was observed in 4 subjects. Cold stress increased the number of isolated pyloric pressurs waves, and decreased the number of antral and propagated antropyloroduodenal pressure waves. Transient phase lll-like duodenal activity occurred in 2 subjects, otherwise there was a reduction in the number of duodenal pressure waves during slress. Basal pyloric pressure was not elevated before, during or after slress. 149

10.5 DTSCUSSTON ln this study, it has been shown that cold stress results in significant modification of pyloric and duodenal motility, as well as the previously demonstrated suppression of antral pressure waves (Stanghellini et al 1983). The characteristic patlern of motility induced by cold stress is similar to that observed in response to other stimuli which also retard the rate of gastric emptying, for example, stimulation of small bowel receptors by dextrose or fat, and infusion of fat into the terminal ileum (Chapter 4). ln addition, the retardant effect of cold slress on gastric emptying was demonstrated to be associated with slowed emptying from the proximal stomach and, in some subjects, with retrograde movement of the meal from the distal to the proximal stomach.

Localized pyloric contractions may retard transpyloric flow (Chapter 6), so it is probable that the increase in IPPWs after cold stress contributed to the delay in gastric emptying, although the substantial suppression of antral and duodenal motility is likely to have been more important. Previously, increased numbers of IPPWs after intraduodenal nutrient infusions have usually been accompanied by tonic elevation of basal pyloric pressurs (Heddle 1988bac). The absence of elevated basal pyloric pressure after cold stress could be due to a difference in stimulus intensity. Although the cold pressor test used in the present study was potent and effective in slowing gastric emplying, local intraduodenal nutrient infusions may result in greater stimulation of pyloric contraction. lnitial episodes of "phase lll-like" duodenal activity were observed in two subjects, both of whom showed evidence of excessive vagal slimulation, with bradycardia in both and hypotension in one. Phase lll-like duodenal activity associated with cold pain has only been previously reported after adrenergic blockade (Thompson et al 1983; Stanghellini et al 1983 a 1984). Duodenal phase lll-like responses have also been observed with labyrinthine stimulation, al levels sufficient to produce nausea (Thompson et al 1982; Stanghellini et al 1983), and after intraduodenal infusions of dextrose, fat and acid (Heddle et al 1988bac; Lewis et al 1979).

Apart from the initial transient increases in duodenal motility in these lwo subjects, there was substantial inhibition of duodenal motility associated with cold stress, a change likely to result in a reduced rate of gastric emptying. This finding is at variance with other studies (Stanghellini et al 1983 & 1984). Possibly this may be due to differences in the position of duodenal pressure sensors. The present results suggest that the intensity of the duodenal responses to stress could be site-specific, with phase lll-like activity observed more distally and inhibition being most marked proximally. Some investigalors did not employ sensors in the duodenal cap (Thompson et al 1982; Stanghellini et al 1983), and although Stanghellini et al (1984) d¡d use proximal duodenal side-holes, it is not clear from their description whether the recordings from those side-holes were analysed separately from sensors located more distally. 150 ln normal individuals, gastric emptying and 'emptying' from the proximal stomach of a digestible solid meal closely approximate a linear pattern (Sheiner et al 1980; Collins et al 1988; Chapter 3). ln the present study, cold stress slowed emptying from both the total stomach, and the proximal gastric region in all subjects. This latter observation, and the retrograde movement of the solid meal which was observed in 4 subjects as gastric emptying slowed, have not been previously reported with cold slress, although similar phenomena were identified after lhe terminal ileal triglyceride infusions described in the previous chapter. Although the alterations in pyloric, antral and duodenal motility observed in this study could partly account for the retrograde movement, changes in regional tone of the stomach, particularly relaxation of the gastric fundus, are also involved in the normal regulation of gastric emptying (Meyer 1987), and could also have contributed to the change in intragastric distribution of food observed after cold stress. The role of neural and hormonal mechanisms in the mediation of the changes in gastric and duodenal motility, gastric emptying and the intragastric distribution of food caused by cold stress is uncertain, although limited information suggesls lhat both factors are of importance (Stanghellini et al 1983 a 1984).

I 0.6 CONCLUSTON The centrally acting stimulus, cold stress, was shown to induce an increase in localized phasic pyloric motility. The association of this effect with the induction of other motor phenomena also observed after small intestinal nutrient infusions, including antral and duodenal suppression, delayed gastric emptying and retrograde movement of solid food, suggests that different stimuli may activate the same final common pathway responsible for a qualitatively similar, coordinated alteration of gastro-pyloro-duodenal motor function (see also Section 9.4). The neurohumoral mediators of these effects, as well as the motor effects of other centrally acting stimuli, such as labyrinthine stimulation, remain unknown. 151

CHAPTER 1 1

Future Manometric Investigation of Pyloric Motor F*unction in Humans

11.1 INTRODUCTION The results of the studies described in this thesis have provided further knowledge about the topography of localized pyloric pressure waves, the pattern of coordinated pyloric motility induced by inlestinal and centrally acting slimuli, and the neurohumoral lransmitters which may stimulate the human pylorus. ln addition, new information has been provided about the interpretation and limitations of intraluminal manometry. lmportantly, although one can still argue about whether the pylorus is the pfimgry agent in the regulation of gastric emptying, a concspt denied vigorously by Thomas (1957) (Section 6.1), the results of these sludies strongly support the belief that the pylorus is of substantial importance in determining the rate of gastric emptying in humans.

This final chapter attempts to summarize the major implications of lhe experiments described in this thesis, and in the light of these, lakes the opportunity of suggesting suitable approaches to the further investigation of human pyloric motor function.

11.2 RECOMMENDATIONS FOR IMPHOVED METHODOLOGICAL APPROACHES FOR PYLORIC MANOMETRY The experiments performed in this thesis have provided new insights into the methodology required for accurate manometric recordings from the pylorus in humans. The experiments described in Chapter I indicate that continuous monitoring of the position of the recording sensor is essential, not only for measuring from the pylorus, but also for the optimal interpretation of antral side-hole manometry. lnformation from the anlral transducer helps to put into clearer 152 context the pressure recordings from sleeve/side-hole manometry. The use ol an antral transducer adds further complexity to studies of pyloric motilily, and may be unnscessary in many situations, as long as the limitations of sleeve/side-hole methods are recognized. ln sleeve assemblies which do not incorporate this device, side-holes spaced at 1 cm intervals along the length of the sleeve sensor should be included to minimize misclassification of localized pyloric pressure waves. In research settings, however, an antral transducer combined with a sleeve/side-hole assembly provides a substantially more complele picture of pyloric motor patterns.

As far as it is possible to determine with currently available techniques, nasoduodenal intubation does not have a significant effect on pyloric motilily. The effects of nasoduodenal intubation, however, may be variable, depending on the level of stress experienced by each subject, or differing degrees of mechanical stimulation of the gastrointestinal mucosa, making demonstration of any effect difficult to show statistically. Possibly, therefore, changes in pyloric motility in response lo nasoduodenal intubation may become evident if grealer numbers of subjects are studied. Further evaluation of this issue may best be tackled by measuring pyloric motor function with and without intubation, using a non-invasive technique such as high resolution ultrasonography in humans (once this becomes technically feasible), or with precisely placed strain gauges in awake dogs. Such an approach may be most appropriate because the initial measuremsnt of pyloric motility without intubation would provide the most suitable control group against which to test the effect of intervention, i.e. intubation.

On the other hand, intubation of the distal small bowel is associated with significanl alterations in antral and pyloric motility. This effect must be taken into account in sludies which use this technique.

11.3 INVESTIGATION OF THE COORDINATED PATTERNS AND PHYSIOLOGICAL CONTROL OF HUMAN PYLORIC MOTILITY The experimsnts reported in Chapter 9 describe the patterns of pyloric motility which occur after infusion of fat into the distal small bowel. Although the modifications of motility observed were qualitatively similar to those observed after intraduodenal infusions of dextrose and triglyceride, the effects of infusion into the ileum of other substances, such as carbohydrate, amino acids and possibly acid, may be different and also warrant evaluation. Recenl studies indicate that the distribution of nutrient/osmotic receptors in the small intestine is site-specific (see Chapter 5), and thus there are likely to be differential effects on motility depending on the site of infusion. Therefore, the effect of jejunal nutrients on the pylorus is also worthy of examination. 153

The question of whether the effect on pyloric, antral and duodenal motility of intraluminal nutr¡ents is mediated by their caloric composition or osmolarity remains unresolved, but may be addressed by sleeve/side-hole manomelric studies comparing the motor effects of nutrient versus non-nutrient infusates of equal osmolarity.

Mechanorecêptors in the small bowel and in the region of the pylorus may well influence pyloric motor function, but this has not been tested. Feedback from mechanoreceptors could be assessed in humans using a variety of stimuli, for example 1) rapid intraduodenal boluses, of increasing volume, of acaloric, isotonic solutions wilh neutral pH; and 2) graded distension at different sites of the duodenum and jejunum with an inlraluminal balloon. lt would also be of interest to study the effect of fundal and antral distension, with acaloric meals of different volumes, or more easily with a barostat balloon. The findings of such studies would not only increase understanding of the factors controlling the pylorus, but may have important clinical implications in patients with gastroparesis, non-ulcer dyspepsia or other disorders in which abdominal bloating and early satiety ars present. Similarly, the existence of the gastrocolic reflex and the known effect of rectal distension on fundal tone and gastric emptying suggest that rectal distension may also influence pyloric motility, and this warrants study.

The patterns of coordinated pyloric motility under conditions of central nervous system stimulation by cold pain were also documented in this thesis. Other central nervous system stimuli are known to slow gastric emptying, and additional studies of the pyloric motor responses to these stimuli are necessary. Acute labyrinthine stimulation, and the possibly more physiological stimulus of psychological stress, could be easily tested. Although logistically more difficult, pyloric motility could also be measured before and after vigorous exercise.

The neural control of the human pylorus has not been previously investigated using appropriate melhodology. The study described in this thesis demonstrated that muscarinic pathways are involved in the pyloric motor response to intraduodenal dextrose. Thus, atropine is a suitable antagonisl against which lo test the pyloric motor responses to other nutrients, or to exogenous administration of candidate hormonal regulators such as CCK. Atropine has recently been demonstrated to inhibit the phasic and tonic pyloric responses to intraduodenal triglyceride (Fraser et al 1989b). This observation suggests that the mechanisms involved in mediating the feedback from small intestinal receptors in humans may be similar to those in dogs. Other mediators found to be relevant in the dog should be evaluated in humans where ethically appropriate. The role of other neurotransmitters should also be investigated using suitable p adrenergic receptor blockers and opiate antagonists such as naloxone. 154

There is little information about the role of gastrointestinal hormones in lhe control of pyloric motility. No hormone has yet been clearly shown to be a regulalor of human pyloric motor function. Recently, inlravenous bolus doses of cholecystokinin octapeptide have been shown to stimulate the pylorus (Fraser et al 1989a). Cholecystokinin and other gastrointestinal hormones need to be administered in doses which are judged to be physiological, but it is difficult to be sure that an intravenous dose reproduces physiological concentrations at the site of action, even if peripheral plasma levels are similar to those found post-prandially. The best approach may be to test the effects of specific hormone antagonists on lhe pyloric motor responses induced both by exogenous hormones and particularly by physiologically appropriate stimuli, such as intraduodenal nutrient infusions or nutrient-rich test meals.

11.4 THE IMPORTANCE OF THE HUMAN PYLORUS AND COORDINATED PYLORIC MOTILITY IN THE REGULATION OF GASTRIC EMPTYING lmportant information about the mechanical role of the pylorus can be obtained from studies using concurrent measurements of motility and transit. Demonstration in this thesis of an inverse correlation between the number of phasic pyloric pressurs waves and the rate of gastric emptying has provided quantitative evidence that the pylorus is important in the regulation of the rate of solid gastric emptying in humans. Further studies with sleeve/side-hole manometry need to be done in humans to measure numbers of IPPWs which occur under other physiological conditions known to be associated with altered rates or differing motor mechanisms of gaslric emptying. For example, pyloric motor patterns during the emptying of meals of different viscosities and densities should be assessed.

Concurrent measurements of proximal gastric tone, with a fundal barostat, and anlropyloroduodenal motility have not been performed. Such studies would provide a substantial technical challenge, but are feasible and are essential to further understanding of "gastroduodenal coordination' in the broadest sense of this term. lnsights could be gained 1) by simultaneous measurement of manometric pressures and fundal tone under a variety of experimental conditions, or 2) by observing any changes in pyloric motility when the variable of gastric lone is controlled by maintaining a constant barostat pressure. Of similar importance is the evaluation of the role of duodenal resistance, which could be addressed by using an intraduodenal barostat. lnformation aboul the importance of fundal tone and duodenal resistance could also be provided by suitable chronic animal models, either by manometry/barostat recordings, or by examining the effect of fundsctomy or proximal duodenal drainage. Such surgical models, however, cannot be constructed without risk of interfering with normal anatomical relationships and nerve supply to the pylorus, so there would always be some doubt about their physiological relevance. 155

The relationship of pressure wave amplitude lo the rate of gastric emptying was considered, at least for the antrum, in the experiments using the antral wall movement detector. lt was shown that stimulation of the manometric pattern of IPPWs was associated with suppresion of antral prêssure waves of any amplitude, as far as the recording technique could determine. Further evaluation of very proximal or very low amplitude antral contractions would require methodology more sensitive than the transducer describêd here. This is unlikely to be achieved by intraluminal techniques, but improvements in non-invasive approaches such ultrasonography or rapid frame scintigraphy may well prove to be satisfactory, while avoiding the degree of radiation exposure involved in lengthy fluoroscopy.

No study has examined the amplitude of IPPWs under differing experimental condilions, nor correlated amplitudes with rates of lranspyloric flow, although the strength of lumen occlusion at the pylorus may well be important. Recording maximum IPPW amplitudes requires relat¡vely low gain settings on the recorder to avoid electronic clipping of the signal, conflicting with the desire to accurately determine the degree of pyloric tone, which requires a high gain because of the very low pressures involved (perhaps only 2 to 5 mmHg). ln most sleeve studies reported so far in humans (Heddle et al 1988a,b,c & 1989; Houghton et al 1988aab), and in the stud¡es in this thesis, high gain settings were preferentially chosen to measure pyloric tone. ln the future, it would be of interest to divide the pyloric signal from the sleeve sensor electronically, so that both high and low gain recordings can be analysed.

Patterns of antropyloroduodenal coordination recorded under these varying experimental circumstances could be assessed by manual inspection of tracings, as was done in this thesis. This method is satisfactory for differentiation of propagated from synchronous pressure waves, but is painstaking when multiple recording sites are employed, and cannot be expected to properly identify the full range and subtle changes of pyloric motor behaviour. Furthermore, unless very rigid scoring criteria are adopted, inter- and intra-observer errors may be substantial, especially during periods of relatively little pressure wave activity (Andersen et al 1989). Patterns of coordination may be more efficiently and accurately evaluated in the future by data digitization and subsequent computer analysis. Some basic programmes for motility analysis are already available for the oesophagus, and the development of hardware and software which can handle the complex patterns generated by the pylorus should be a priority.

This thesis has not attempted to address issues of abnormal pyloric function. However, improved understanding of the demands of manometry and the normal patterns of pyloric molility enables more adequate evaluation of disordered motility and the relevance of pyloric dysfunction to abnormal gastric emptying. Clinical conditions requiring investigation include gastroparesis especially in diabetic patients, non-ulcer dyspepsia, post-vagotomy and/or pyloroplasty, and 156

probably peptic ulcer disease. There is some preliminary evidence lhat pyloric dysfunction may occur in diabetic gastroparesis (Mearin et al 1985), but this needs more formal evaluation. Certainly in these patients there are almost certain to be abnormalities of neural innervation as well as some degree of myopathy, so pyloric motility is likely to be affected. The major problem with the manometric investigation of these patients is that abnormalities of antral peristalsis may make transpyloric intubation more difficult or impossible in some patients, unless fluoroscopy or endoscopy is initially used lo help position the recording assembly. Conceivably, also, abnormalities of mucosal innervation, acid output or cellular metaplasia may aller anlral or duodenal TMPD measurements, making the use of this technique to monitor assembly position throughout the study more difficult lo interpret.

11.s CONCLUSTON Only in recent years has suitable melhodology been developed to allow accurate pyloric manometry. The studies described in this thesis shed new light on several aspects of the control and function of the human pylorus. The ability to record human pyloric motility concurrent with measuremenls of gastric emplying, together with the knowledge required to interpret these recordings properly, provide a basis which is essential for increased understanding of the mechanical role of the pylorus and its relationship to global gastroduodenal motor function. 157

Appendix

PUBLICAT¡ONS RESULTING FROM THE STUDIES REPORTED IN THIS THESIS

Sclentlf lc Papers

Fone DR, Horowitz M, Dent J, Read NW, Heddle R. Pyloric motor response to intraduodenal dextrose involves muscarinic mechanisms. Gastroenterology 1989; 97: 83-90.

2 Fone DR, Horowitz M, Maddox A, Akkermans LMA, Read NW, Dent J. Gastroduodenal molility during the delayed gastric emptying induced by cold stress. Gastroenterology

1 990; 98: 1 1 5s-1 1 61 .

3 Fone DR, Horowitz M, Read NW, Dent J, Maddox A. The effect of terminal ileal triglyceride infusion on gastroduodenal motility and the intragastric distribution of a solid meal. Gastroenterology 1990; 98: 568-575.

4 Fone DR, Akkermans LMA, Dent J, Horowitz M, Van der Schee E. Evaluation of patterns of human antral and pyloric motility with an antral wall motion detector. Am. J. Physiol. 1990; 258 (Gastrointest. Liver Physiol. 21): G616-G623.

5 Fone DR, Horowitz M, Heddle R, Maddox AF, Collins PJ, Read NW, Dent J. Comparative effects of duodenal and ileal intubation on gastric emptying and postprandial antral, pyloric and duodenal motility. Scand. J. Gaslroenterol. 1990 (in press).

Ab st ract s

Fone D, Heddle R, Horowitz M, Maddox A, Collins P, Dent J. Comparative effects of duodenal and ileal intubation on gastric emptying and antropyloroduodenal motility. Gastroenterology 1988; 94: A133. 158

2 Fone D, Maddox A, Horowitz M, Read N, Dent J, Collins P. Antropyloroduodenal motor responses to terminal ileal lipid infusion - motor correlates of the ileal brake. Gastroenterology 1988; 94: 4133.

3 Fone D, Horowitz M, Maddox A, Akkermans L, Denl J, Read N. Cold stress induces isolated pyloric pressure waves in healthy volunleers. Gastroenterology 1988; 94 A1 33.

4 Fone D, Akkermans L, Dent J, Horowitz M, Van der Schee E. Are isolated pyloric pr€ssure waves really isolated? Gastroenterology 1988; 94: 44.

5 Fone D, Heddle R, Horowitz M, Dent J, Read N. Atropine blocks the pyloric motor response to intraduodenal dextrose in healthy volunteers. Gastroenterology 1988; 94: 4132.

6 Fone D, Maddox A, Horowitz M, Read N, Dent J, Collins P. Antropyloroduodenal motor responses to terminal ileal lipid infusion. Aust. NZ. J. Med. 1988; 18:457. 159

Bibltography

Abell TL, Malagelada J-R, Lucas AR, Camilleri M, Go VLW, Azpiroz F, Callaway CW, Kao PC, Zinsmeister AR, Huse DM 1987. Gastric electromechanical and neurohumoral function in anorexia nervosa. Gastroenterology 93: 958-965.

Adam RD, Heading RC, Anderson T, McDicken WN 1982. Application of real-time ultrasonic imaging to the study of gastric motility. ln: Weinbeck M Ed., Motility of the Digestive Tract. Raven Press, New York, pp. 215-218.

Adams PH, Davies RE 1963. Source of the gastric mucosal potential difference. Nature 197: 874-876.

Adams RG, Verma P, Jackson AJ, Miller RL 1982. Plasma pharmacokinetics of intravenously administered atropine in normal human subjects. J. Clin. Pharmacol.22:477-481.

Adrian T, Savage A, Sagor G, Allen J, Baracese-Hamilton A, Tatemoto K, Polak J, Bloom S 1985. Effect of peptide YY on gastric, pancreatic and biliary funclion in humans. Gastroenterology 89: 494-499.

Aeberhard P, Bedi BS 1977. Effects of proximal gastric vagolomy (PGV) followed by total vagotomy (TV) on postprandial and fasting myoelectrical activity of the canine stomach and duodenum. Gut 18: 515-523.

Agostoni E, Chinnock JE, De Burgh Daly M, Murray JG 1975. Functional and histological studies of the vagus nerve and its branches to the heart, lungs and abdominal viscera in the cat. J. Physiol. (Lond.) 135: 182-2os.

Akkermans LMA, Jacobs F, Hong-Yoe O, Roelofs JMM, Wittebol P 1980. A non-invasive method to quantify antral contractile activity in man and dog (a preliminary report). ln: Christensen J Ed., Gastrointestinal Motility. Raven Press, New York, pp. 195-202. 160

Allen GL, Poole EW, Code CF 1964. Relationship between electrical activities of antrum and duodenum. Am. J. Physiol. 207: 906-910.

Allen JM, Fitzpatrick ML, Yeats JC, Darcy K, Adrian TE, Bloom SR 1984. Effects of peptide YY and neuropeptide Y on gastric emptying in man. Digestion sOi 255-262.

Allescher H-D, Daniel EE, Dent J, Kostolanska F, Fox JET 1987. Cholinergic reflex activation of canine pyloric motor activity from the proximal duodenum in vivo. Dig. Dis. Sci. 32(8): 900.

Allescher H-D, Ahmad S, Daniel EE, Dent J, Kostolanska F, Fox JET 1988a. lnhibitory opioid receptors in canine pylorus. Am. J. Physiol.255 (Gastrointest. Liver Physiol. 18): G352-G360.

Allescher H-D, Daniel EE, Dent J, Fox JET, Kostolanska F 1988b. Extrinsic and intrinsic control of pyloric sphincter pressure in the dog. J. Physiol. (Lond.) 401:17-38.

Allescher H, Dent J, Fox JE, Daniel EE 1988c. Control of canine pylorus function by neuropeptides (unpublished).

Allescher HD, Daniel EE, Dent J, Fox JET 1989a. lnhibitory function of VIP-PHI and galanin in canine pylorus. Am. J. Physiol.256 (Gastrointest. Liver Physiol. 19): G789-G797.

Allescher HD, Dent J, Fox JET, Daniel EE, Classen M 1989b. Modulatiort of canine pyloric motor activity via adrenergic receptors. J. Pharmacol. Exper. Ther. 249: 652-659..

Allescher H-D, Daniel EE, Dent J, Fox JET, Kostolanska F 1989c. Neural reflex of the canine pylorus to intraduodenal acid infusion. Gastroenterology 96: 18-28.

Alumets J, Fahrenkrug J, Hakansson R, Schaffalitzky de Muckadell OB, Oddman R 1979. A rich VIP nerve supply is characteristic of sphincters. Nature 280: 155-156.

Alvarez WC 1940. The movements of the stomach. ln: Alvarez WC Ed., An lntroduction to Gastroenterology, Third Edition. Hueber lnc., New York Publishers, pp. 287-324.

Alvarez WC, Mahoney 1922. Action currenls in stomach and intestine. Am. J. Physiol. 58: 476- 493.

Amidon GL 1985. Fluid mechanics and intestinal transit. Gastroenterology 88:858. 161

Andersen MB, Stodkilde-Jorgensen H, Kraglund K, Djurhuus JC, Rosenfalck A 1989. Consistency in manual scoring analysis of gastrointestinal pressure recordings. Scand. J. Gastroenterol. 24: 321-328.

Andersson S, Grossman Ml 1965. Profile of pH, pressure and potential difference at the gastroduodenal junction in man. Gastroenterology 49: 364-371.

Anuras S, Cooke AR, Christensen J 1974. An inhibitory innervation at the gastroduodenal junction. J. Clin. lnvest. 54: 529-535.

Arndorfer RC, Stef JJ, Dodds WJ, Lineham JH, Hogan WJ 1977. lmproved infusion system for intraluminal esophageal manometry. Gastroenterology 73: 23-27.

Aste H, Pandolfo N, Pugliese V, Nebiacolombo C 1979. Prolonged endoscopic and manometric observations of the pylorus. Scand. J. Gastroenterol 14 (Suppl. 54\:72-77.

Atkinson M, Edwards DAW, Honour AJ, Rowlands EN 1957. Comparison of cardiac and pyloric sphincters. A manometric study. Lancet ii: 918-922.

Avill R, Mangnall YF, Bird NC, Brown BH, Barber DC, SehgarAD, Johnson AG, Read NW 1987. A new non-invasive lechnique for measuring gastric empty¡ng. Gastroenterology 92: 1019-1026.

Azpiroz F, Malagelada J-R 1985a. lntestinal control of gastric tone. Am. J. Physiol. 249 (Gastrointest. Liver Physiol. 12): G501-G509.

Azpiroz F, Malagelada J-R 1985b. Physiological variations in canine gastric tone measured by an electronic barostat. Am. J. Physiol.248 (Gaslrointest. Liver Physiol. 11): G229-G237.

Barajas-Lopez C, Berezin l, Daniel EE, Huizinga JD 1989 Pacemaker aclivity recorded in interstitial cells of Cajal of lhe gastrointestinal tract. Am. J Physiol. 257 (Cell Physiol. 26): c830-c835.

Barbezat GO, Grossman Ml 1979. Cholecystokinin released by duodenal acidificalion. Gastroenterology 60: 761.

Bass P, Code CF, Lambert EH 1961. Electric activity of gastroduodenal junction. Am. J. Physiol.; 2O1: 587-592. 162

Bateman DN, Whittington TA 1982. Measurement of gastric emptying by real-time i,i ultrasonography. Gut 23; 524-527.

Bedi BS, Kelly KA, Holley KE 1972. Pathways of propagation of the canine gastric pacesetter potential. Gastroenterology 63: 288-296.

Behar J, Biancani P, Zabinski MP 1979. Characterization of feline gaslroduodenal junction by neural and hormonal stimulation. Am. J. Physiol. 236: E45-E51.

Berk T, Crochelt RF, Peikin SR 1986. Role of pylorus in mediating cholecystokinin-stimulated satiety in the Zucker rat. Dig. Dis. Sci. 31: 502-505.

Bertiger G, Reynolds JC, Ouyang A, Cohen S 1987. Properlies of the feline pyloric sphincter in vitro. Gastroenterology 92: 1965-1972.

Biancani P, Zabinski MP, Kerstein MD, Behar J 1980. Mechanical characleristics of the cat pylorus. Gastroenlerology 78: 301-309.

Biancani P, CiOalzi LK, McCallum RW 1981. Mechanism of histamine-induced excitation of the cat pylorus. J. Clin. lnvest. 68: 582-588.

Bitar K, Mei N, Michelucci MH 1975. Vagal mechanoreceptors of the LOS and of the pyloric sphincter in the cat. J. Physiol. (Lond.) 245: 103-104P.

Bortoff A 1990. lntramural control of gastro-duodenal coordination. ln: Van Neuten, Schuurkes and Akkermans Eds., Gastro-pyloro-duodenal Coordination. Wrightson Biomedical Publishing Ltd., Petersfield. pp. 83-96.

Bortoff A, Davis RS 1968. Myogenic transmission of antral slow waves across lhe gastroduodenal junction in situ. Am. J. Physiol. 215: 889-897.

Bortoff A, Weg N 1965. Transmission of electrical activity lhrough lhe gastroduodenal junction Am. J. Physiol. 208: 531-536.

Bovell KT, Tougas G, Dent J, Collins SM, Hunt RH 1987. Effect of naloxone on lipid induced pyloric motor responses in humans. Gastroenterology 92: 1325. 163

Brener W, Hendrix TR, McHugh PR 1983. Regulation of the gastric emptying of glucose Gastroenterology 85: 76-82.

Brink BM, Schlegel JF, Code CF 1965. The pressure profile of the gastroduodenal junctional zone in dogs. Gut 6: 163-171.

Brody DA, Quigley JP 1944. Application of the 'inductograph" to the registration of movements of body structures such as the pyloric sphincter. J. Lab. Clin. Med. 29: 863-867.

Brody DA, Werle JM, Meschan l, Quigley JP 1940. lntralumen prsssurês of the digestive tract, especially the pyloric region. Am. J. Physiol. 130: 791-801.

tttl-HSA Bromoten D, Carlberger G, Lundh G 1968. Measurement of gastric emptying rate using Scand. J. Gastroenterol. 3: 641-653.

Brown JC, Dryburgh JR, Ross SA, Dupre J 1975. ldentification and actions of gastric inhibitory polypeptide. Rec. Prog. Horm. Res. 31: 487-532.

] rd Brunemeir EH, Carlson AJ 1915. Contributions to the physiology of the stomach. XIX: reflexes .i: ','! I from the intestinal mucosa to the stomach. Am. J. Physiol.36:191-195.

Bueno L, Fioramonti J, Ferre JP 1983. Calcitonin - CNS action to control the pattern of intestinal motilily in rats. Peptides 4: 63-66.

Cai and Gabella 1984. Structure and innervation of the musculature at the gastroduodenal junction of the guinea pig. J. Anat. 139: 93-104.

Camilleri M, Malagelada J-R 1984. Abnormal intestinal motility in diabetics with the gaslroparesis syndrome. Eur. J. Clin. lnvest. 14:. 42O-427.

Camilleri M, Malagelada J-R, Brown ML, Becker G, Zinsmeister AR 1985. Relation between antral motility and gastric emptying of solids and liquids in humans. Am. J. Physiol. 249 (Gastrointest. Liver Physiol. 12): G580-G585.

Camilleri M, Malagelada JR, Stanghellini V, Zinsmeister AR, Kao PL, L¡ CH 1986. Dose-related I etfects of synthetic human ß-endorphin and naloxone on fed gastrointestinal motility. Am. J. Physiol. 251 (Gastrointest. Liver Physiol. 14): G147-G154. t 164

Camilleri M, Colemont LJ, Phillips SF, Brown ML, Thomsforde GM, Chapman N, Zinsmeister AR 1989. Human gastric emptying and colonic filling of solids characterized by a new method. Am. J. Physiol. 257 (Gastrointest. Liver Physiol. 20): G284-G290.

Cann PA, Read NW, Cammack J, Childs H, Holden S, Kashman R, Longmore J, Nix S, Simms N, Swallow K, Weller J 1983. Psychological stress and the passage of a standard meal through the stomach and small intestine in man. Gul24: 236-240.

Cannon WB 1898. The movements of the stomach studied by means of the rontgen rays. Am. J. Physiol. 1: 359-382.

Cannon WB 1907. The acid control of the pylorus. Am. J. Physiol. 2O: 283-322.

Cannon WB, Lieb CW 1911. The receptive relaxation of the stomach. Am. J. Physiol. xxüi:267- 273.

Carlson AJ, Litt S 1924. Studies on the visceral nervous system and on the reflex control of the pylorus. Arch. lnt. Med. 33: 281-291.

-r

Carlson HC, Code CF, Nelson RA 1966. Motor action of the canine gastroduodenal junction: a ', cineradiographic, pressure, and electric study. Am. J. Dig. Dis. 11:155-172.

Chaussade S, Redah D, Nepveux P, Couturier D, Guerre J 1989. Antral (A), pyloric (P), bulb (B) and duodenal (D) motor contributions to duodenogastric reflux (DGR) in humans. Gastroenterology 96: 484 (abstract).

Chey WY, Hitanant S, Hendricks J, Lorber SH 1970. Effect of secretin and cholecystokinin on gastric emptying and gastric secretion in man. Gastroenterology 58: 820-827.

Christensen J 1987. Motility of the Colon. ln Johnson LR Ed. Physiology of the Gastrointestinal Tract Volume 1. New York, Raven press: pp665-693.

Christian PE, Moore JG, Sorenson JA et al 1980. Effects of meal size and correction lechnique on gastric emptying time: studies wilh lwo tracers and opposed detectors. J. Nucl. Med. 21: 883- 88s. I

! 165

Christian PE, Dalz FL, Sorenson JA, Taylor A 1983. Technical factors in gastric emptying studies. J. Nucl. Med. 24(3): 264-268.

Code CF 1970. The mystique of the gastroduodenal junction. Rendiconti di Gastroenterologica 2 20-37.

Code CF, Szurszewski JH, Kelly KA, Smith lB 1968. A concept of control of gastrointestinal motility. ln: Code CF Ed., Handbook of Physiology, Volume 5, Section 6. Am. Physiol. Soc., Bethesda, Maryland, pp. 2881-2896.

Code CF, Schlegel JF 1974. The gastrointestinal housekeeper. ln: Daniell EE Ed., Gastrointestinal Motility. Mitchell Press, Vancouver, pp. 631-633.

Code CF, Steinbach JH, Schlegel JF, Amberg JA. Hallenbeck GA 1984. Pyloric and duodenal motor contributions to duodenogastric reflux. Scand. J. Gastroenterol. 19 (Suppl.92):13-16.

Cole LG 1917. The motor phenomena of the stomach, pylorus, and cap observed roentgenographically Am. J. Physiol. 42: 618. .ì q j Collins PJ, Horowitz M, Cook DJ, Harding PE, Shearman DJC 1983. Gastric emplying in normal subjects - a reproducible lechnique using a single scintillation camera and computer system. Gut 24:1117-1125.

Collins PJ, Horowitz M, Shearman DJC, Chatterton BE 1984. Correction for tissue attenuation in radionuclide gastric emptying studies: a comparison of a lateral image method and a geometric mean method. Br. J. Radiol. 57: 689-695.

Collins PJ, Horowitz M, Chatterton BE 1988. Proximal, distal and total stomach emptying of a digestible solid meal in normal subjects. Br. J. Radiol.61: 12-18.

Cook MA, Kowalewski K, Daniel EE 1977. The use of the isolated, homologous perfused canine stomach in the study of the loci and mechanism of action of gastrointestinal hormones. Materia

Medica Polona 9: 1 13-1 1 7. t I Cooke AR 1974. Duodenal acidification: role of the first part of the duodenum in gastric emptying ; and secretion in dogs. Gastroenterology 67: 85-92.

l 166

Cooke AR 1975. Control of gastric emptying and motility. Gastroenterology 68: 804-816

Cooke AR 1977. Localization of receptors inhibiting gastric emptying in the gul. Gastroenterology 72; 875-880.

Cooke AR 1978. Gastric emptying in the cat in response to hypertonic solutions and tryptophan Dig. Dis. 23: 312-315.

Cooke AR, Kotteman WJ 1974. Antral and duodenal motor activity and gaslric emptying with secretin and pentagastrin. ln: Daniel EE Ed., Proceedings of the 41h lnlernational Symposium on Gastroinlestinal Motility. Mitchell Press, Vancouver, pp. 537-545.

Cortot A, Phillips SF, Malagelada J-R 1982. Parallel gastric emptying of nonhydrolyzable fat and water after a solid-liquid meal in humans. Gastroenterology 82: 877-881.

Cottrell DF, lggo A 1984. Tension receptors wilh vagal afferent fibres in the proximal duodenum and pyloric sphincter of sheep. J. Physiol. (Lond.) 354: 457-475.

Crider JO, Thomas JE 1937. A study of gastric emptying with the pylorus open. Am. J. Dig. Dis 4: 295-300.

Daniel EE 1965. The electrical and contractile activity of the pyloric region in dogs and the effects of drugs. Gastroenterology 49: 403-418.

Daniel EE and Allescher H-D 1990. Structure of the pyloric region. ln: Van Neuten, Schuurkes and Akkermans Eds., Gastro-pyloro-duodenal Coordination. Wrightson Biomedical Publishing Ltd., Petersfield, pp. 19-34.

Daniel EE, Posey-Daniel V 1984a. Neuromuscular structures in oppossum esophagus. Role of interstitial cells of Cajal. Am. J. Physiol. 246: G305-G315.

Daniel EE, Posey-Daniel V 1984b. Effects of scorpion venom on slructure and function of esophageal lower sphincter (LES) and body circular muscle (BCM) from opossum. Can. J. Physiol. Pharmacol. 62: 360-373. I I

I Daniel EE and Sarna S 1978. The generation and conduction of electrical activity in smooth muscle. Ann. Rev. Pharmacol. Toxicol. 18: 145-166.

þ 167

Daniel EE, Wiebe GE 1966. Transmission of reflexes arising on both sides of the gastroduodenal junction. Am. J. Physiol. 211: 634-642.

Daniel EE, Carlow DR, Wachter BT, Sutherland WH, Bogoch A 1959. Electrical activity of the small intestine. Am. J. Physiol. 37: 268-281.

Daniel EE, Berezin l, Allescher H-D, Manaka H, Posey-Daniel VP 1989. Morphology of the canine pyloric sphincter in relation to function. Personal communication.

Debas MT, Farooq O, Grossman Ml 1975. lnhibition of gastric emptying is a physiological aclion of cholecystokinin. Gastroenterology 68: 121'l-1217.

Defilippi CC, Gomez E 1985. Continuous recording of pyloric sphincter pressure in dogs. Relationship to migratory motor complex. Dig. Dis. Sci. 30: 669-674.

Deloof S 1988. Sympathetic control of antral and pyloric electrical activity in the rabbit. J. Autonomic Nerv. Syst.; 22: 1-1o.

Deloof S, Rousseau JP 1985. Specific effects of thoracic vagotomy on the electrical activity of the gastric antrum and pylorus in rabbils. Quart. J. Exp. Physiol. 70: 491-501.

Dent J 1976. A new technique for continuous sphincter pressure measurement. Gastroenterology 71:263-267.

Diamant NE, Hall K, Mui H, El-Sharkawy TY 1980. Vagal control of the feeding motor pattern in the lower esophageal sphincter, stomach and small intestine of dog. ln: Christensen J Ed., Gastrointestinal Motility. New York, Raven Press, pp. 365-370.

Dockray GJ 1988. Regulatory peptides and the neuroendocrinology of gut-brain relations. Quart. J. Exp. Physiol. 73: 703-727.

Dodds WJ 1976. lnslrumentation and methods for intraluminal esophageal manometry. Arch lntern. Med. 136: 515-523.

Dooley CP, Reznick JB, Valenzuela JE 1984. Variations in gastric and duodenal motility during gastric emptying of liquid meals in humans. Gastroenterology 70: 117-123. 168

Dooley CP, Reznick JB, Valenzuela JE 1985. A continuous manometric study of the human pylorus. Gastroenterology 89: 821-826.

Dozois RR, Kelly LA, Code CF 1971. Effect of distal antrectomy on gastric emptying of liquids and solids. Gastroentreology 61 : 675-681.

Dugas MC, Schade RR, Lhotsky D, Van Thiel D 1982. Comparison of methods for analysing gastric emptying. Am. J. Physiol. 243: G237-G242.

Duthie HL, Kwong NK, Brown BH, Whittaker GE 1971. Pacesettter polential of the human gastroduodenal junction. Gut 12: 250-256.

Edin R 1980. The vagal control of the pyloric motor function. A physiological and immunohistochemical study in cat and man. Acta. Physiol. Scand. Suppl.485: 1-30.

Edin R, Ahlman H, Kewenter J 1979a. The vagal control of the feline pyloric sphincter. Acta. Physiol. Scand. 107: 169-174.

Edin R, Lundberg JM, Ahlman H, Dahlstrom A, Fahrenkrug J, Hokfelt T, Kewenter J 1979b. On the VIP-ergic innervation of the feline pylorus. Acta. Physiol. Scand. 107:185-187.

Edwards DAW, Rowlands EN 1968. Physiology of the gastroduodenal junction. ln: Code CF Ed., Handbook of Physiology - Alimentary Canal lV. Waverly Press, Baltimore Pub., pp. 1985-2000.

Ehrlein H-J 1980. A new technique for simultaneous radiography and recording of gastro¡ntestinal motilily in unanaesthetized dogs. Lab. Anim. Sci. 30: 879-884.

Ehrlein H-J, Prove J 1982. Effect of viscosity of tost meals on gastric emptying in dogs. Quart. J. Exp. Physiol. 67: 419-425.

Ehrlein H-J 1988. Motility of the pyloric sphincter studied by the inductograph method in conscious dogs. Am. J. Physiol.254 (Gastrointest. Liver Physiol. 17): G650-G657.

Ehrlein HJ, Hiesenger E 1982. Computer analysis of mechanical activity of gaslroduodenal junction in unanaesthetized dogs. Quart. J. Exp. Physiol. 67: 17-29. 169

Elashoff JD, Reedy TJ, Meyer JH 1982. Analysis of gastric emptying data. Gastroenterology 83

1 306-1 31 2.

Elias E, Gibson GJ, Greenwood LF, Hunt JN, Tripp JH 1968. The slowing of gastric emptying by monosaccharides and disaccharides in test meals. J. Physiol. 194: 317-326.

Elfvin L-G, Lindh B 1982. A study of the extrinsic innervation of the guinea pig pylorus with the horseradish peroxidase tracing technique. J. Comp. Neurology 208: 317-324.

El-Sharkawy TY, Szurszewski JH 1978. Modulation of canine antral circular smooth muscle by acetylcholine, noradrenaline and penlagastrin. J. Physiol. 279: 309-320.

El-Sharkawy TY, Morgan KG, Szurszewski JH 1978. lntracellular electrical activity of canine and human gaslric smooth muscle. J. Physiol. 279:291-3O7.

Ewart WR, Wingate DL 1984. Central representation of arrival of nulrient in the duodenum. Am. J. Physiol. 246: G75o-G756.

Eyre-Brook lA, Smallwood RH, Linhardt GE, Johnson AG 1983. Timing of pyloric closure in man Studies with impedance electrodes. Dig. Dis. Sci. 28:. 1106-1115.

Eysenck HJ. Manual of the Maudsley Personality Inventory. University of London Press, London.

Faichney A, Chey W, Kim Y, Lee K, Kim M, Chang T 1981. Effect of sodium oleate on plasma secretin concenlration and pancreatic secretion in dog. Gastroenlerology 81'. 458-462.

Feldman M, Smith HJ, Simon TR 1984. Gastric emptying of solid radiopaque markers: studies in healthy subjects and diabetic patients. Gastroenterology 87: 895-902.

Fisher R, Cohen S 1973. Physiological characteristics of the human pyloric sphincter. Gastroenterology; 64: 67-75.

Fisher RS, Lipshutz W, Cohen S 1973. The hormonal regulation of pyloric sphincter function. J. Clin. lnvest. 52: 1289-1296.

Fraser R, Fone D, Horowitz M, Dent J 1989a. Effects of cholecystokinin octapeptide on anlral, pyloric and duodenal motility. Aust. NZ J. Med. 19 (suppl.): 593 (abstract). 170

Fraser R, Fone D, Horowitz M, Dent J 1989b. Pyloric molor responsê to intraduodenal lipid is sustained and atropine sensitive. Gastroenterology 86: 4158 (abstract).

Fried M, Lochner C, Erlacher U, Beglinger C, Jansen J, Lamers C, Stalder GA 1989. Role of CCK in the regulation of gastric emptying and pancreatic secretion in man. Gastroenterology 96:4159 (abslracl).

Furness JB, Costa M 1987. Sympathetic influences on gastrointestinal function. ln: The Enteric Nervous System, Chapter 9. Churchill Livingstone Pub., pp. 2O7-238.

Fujii Y 1971. Electrophysiological studies on gastroduodenal junction of guinea pig. Am. J Physiol. 221: 413-420.

Gabella G 1972. Fine structure of the myenteric plexus in the guinea pig ileum. J. Anat. 11 1: 69- 97.

Gabella G 1987. Structure of muscles and nerves in the gastrointestinal lract. ln: Johnson LR Ed., Physiology of the Gastrointestinal Tracl, 2nd Edition, Volume 1. Raven Press, New York, pp. 335-382.

Gaffney PR, Gleeson DJ, Hall JW, Brady MP 1987. The manometric findings at the human pylorus. Scand. J. Gastroenlerol. 22: 525-532.

Geall MG, Phillips SF, Summerskill WHJ 1970. Profile of gastric potential difference in man. Gaslroenterology 58: 437-443.

Geldof H, van der Schee EJ, van Blankenstein M, Grashuis JL 1986. An electrogastrographic study of gastric myoelectric activity in patients with unexplained nausea and vomiting. Gul 27: 799-808.

George JD 1968. New clinical method for measuring the rate of gastric emptying: the double sampling test meal. Gut 9: 237-242.

Gerner T, Haffner JFW 1978. The role of local cholinergic pathways in the motor response to cholecystokinin and gastrin in isolated guinea pig fundus and anlrum. Scand. J. Gastroenterol. 12:

7 51 -7s7 . 171

Gershon-Cohen J, Shay H 1937. Experimental studies on gastric physiology in man lll. A study of pyloric control. The role of milk and cream in the normal and in subjects with quiescent duodenal ulcer. Am. J. Roentgenol. 38: 427-446.

Gilbey SG, Watkins PJ 1987. Measurement by epigastric impedance of gastric emptying in diabetic autonomic neuropathy. Diab. Med. 4: 122-126.

Goldstein H, Boyle JD 1965. The saline load test - a bedside evaluation of gastric retention Gastroenterology 49: 375-380.

Golenhofen K, Ludtke FE, Milenov K, Siewert R 1980. Excitatory and inhibitory effects on canine pyloric musculature. ln: Christensen J Ed., Gastrointestinal Motility. Raven Press, New York, pp. 203-210.

Gonda T, Daniel EE, McDonald TJ, Fox JET, Brooks BD, Oki M 1990. The distribution and funclion of enteric GAL-lR nerves in dogs: comparison with VlP. Am. J. Physiol. 256: G884-G896.

Goo RH, Moore JG, Greenberg E, Alazraki NP 1987. Circadian variation in gastric emptying of meals in humans. Gastroenterology 93: 515-518.

Goodman EN, Colcher H, Kalz GM, Dangler CL 1955. The clinical significance of the electrogastrogram. Gastroenterology 29: 598-607.

Gray H 1989. The vagus nêrve. ln: Williams PL, Warwick R, Dyson M, Bannister LH Eds., Grays Anatomy, 37th edition, Section 7, Neurology. pp. 1113-1118.

Griffith GH, Owen GM, Kirkman S, Shields R 1966. Measurement of rate of gastric emptying using chromium-51. Lancet i: 1244-1245.

Grossman M 1974. Candidate hormones of the gut, lntroduct¡on. Gastroenterology 67: 730-755

Gulsrud PD, Taylor lC, Watt HD, Cohen MB, Elashoff J, Meyer JH 1980. How gastric emptying of carbohydrate affects glucose tolerance and symptoms after truncal vagotomy and pyloroplasty. Gastroenterology 78: 1463-1471.

Hall WH, Read RC 1970. Effect of vagotomy on gastric emptying. Am. J. Dig. Dis. 15: 1047-

1 053. 172

Hara YM, Kubota M, Szurszewski JH 1986. Electrophysiology of smooth muscle of the small intestine of some mammals. J. Physiol. (Lond.) 372:5O1-52O.

Harvey RF, Mackie DB, Brown NJG 1970. Measuremont of gastric emptying time wilh a gamma camera. Lancet i: 16-18.

Hazarika NH, Coote J, Downman CBB 1964. Gastroinlestinal dorsal root viscerotomies in the cal J. Neurophysiol.: 109-1 16.

Heading RC 1980. Gastric motility. Front. Gastrointest. Res. 6: 35-56.

113m Heading RC, Tothill P, Laidlaw AJ, Shearman DJC 1971. An evaluation of lndium chelate in the measurement of gastric emptying by scintiscanning. Gut 12: 611-615.

Heading RC, Tothill P, Mcloughlin GP, Shearman DJC 1976. Gastric emptying rate measurement in man. A double isotope scanning technique for simultaneous study of liquid and solid componenls of a meal. Gastroenterology 71: 45-50.

Heddle R 1989. Function and role of the human pylorus. Doctoral Thesis.

Heddle R, Dent J, Read NW, Houghton LA, Toouli J, Horowitz M, Maddern GJ, Downton J 1988a. Antropyloroduodenal motor responses lo intraduodenal lipid infusion in healthy volunteers. Am. J. Physiol. 254: G671-G679.

Heddle R, Dent J, Toouli J, Read NW 1988b. Topography and measuremenl of pyloric pressure waves and tone in humans. Am. J. Physiol.255 (Gaslrointest. Liver Physiol. 18): G490-G497.

Heddle R, Fone D, Dent J, Horowitz M 1988c. Stimulation of pyloric motility by intraduodenal dextrose in normal subjects. Gul 29: 1349-1357.

Heddle R, Collins PJ, Dent J, Horowitz M, Read NW, Chatterton B, Houghton LA 1989. The motor mechanisms associated with slovrling of the gastric emptying of a solid meal by an intraduodenal lipid infusion. J. Gastroenterol. Hepatol. 4: 437-447.

Hinder RA, Kelly KA 1977. Human gastric pacesetter potential. Site of origin, spread, and response to gastric transection and proximal gastric vagotomy. Am. J. Surg. 133:29-33. 173

Hinder RA, Bremner CG 1978. Relative role of pyloroplasty size, truncal vagotomy, and milk meal volume on canine gastric emptying. Dig. Dis. 23: 210-216.

Hinder RA, San-Garde BA 1983. lndividual and combined roles of the pylorus and the antrum in the canine gastric emptying of a liquid and a digestible solid meal. Gastroenterology 84: 281- 286.

Holgate AM, Read NW 1985. Effect of ileal infusion of intralipid on gastrointestinal lransit, ileal flow rate, and carbohydrate absorption in humans after ingestion of a liquid meal. Gastroenterology 88: 1005-11.

Holt S, McDicken WN, Anderson T, Stewart lC, Heading RC 1980. Dynamic imaging of the stomach by real time ultrasound - a method for the study of gastric motility. Gut 21:597-601.

Hopkins A 1966. The pattern of gastric emptying: a new view of old results. J. Physiol. (Lond.) 182:144-149.

Horowilz M, Akkermans LMA 1989. Scintigraphic measurement of gastric emptying. ln: Read NW Ed., Gastrointestinal Motility: Which Test? Wrighton Biomedical Publishing Ltd, Petersfield, pp. 73-89.

Horowilz M, Cook DJ, Collins PJ, Harding PE, Shearman DJC 1982. The applicalion of techniques using radionuclides to the study of gastric emptying. Surg. Gynaecol. Obstet. 155:737-744.

Horowitz M, Collins PJ, Shearman DJC 1985. Disorders of gastric emptying in humans, and the use of radionuclide techniques. Arch. lnt. Med. 145: 1467-1475.

Horton BT 1928. Pyloric musculature with special reference to pyloric block. Am. J. Anat.41 197 -226.

Horton BT 1931. Pyloric block, with special reference to the musculature, myenteric plexus and lymphatic vessels. Arch. Surg. 22: 438-462.

Houghton LA, Read NW, Heddle R, Horowitz M, Collins PJ, Chatterton B, Dent J 1988a. Relationship of the motor activity of the antrum, pylorus, and duodenum lo gastric emptying of a solid-liquid mixed meal. Gastroenterology 94: 1285-1291. 174

Houghton LA, Read NW, Heddle R, Maddern GJ, Downton J, Toouli J, Dent J 1988b. Motor activity of the gastric antrum, pylorus, and duodenum under fasted conditions and after a liquid meal. Gastroenterology 94: 1276-1284.

Hunt JN 1956. Some properties of an alimentary osmoreceptor mechanism. J. Physiol. 132: 267 -288.

Hunt JN, Knox MT 1968. Regulation of gastric emptying. Chapter 94 ln: Code CF Ed., Handbook of Physiology Section 6 - Alimentary Canal, Volume 4. American Physiological Society, Bethesda, Maryland, pp. 1917-1935.

Hunt JN, Knox MT 1969. The slowing of gastric emptying by 9 acids. J. Physiol. (Lond.) 201 161-179.

Hunt JN, Knox MT, Oginski A 1965. The effect of gravíty on gastric emptying with various test meals. J. Physiol. (Lond.) 178:92-97.

Hunt JN, Spurrell WR 1951. The pattern of emptying of the human stomach. J. Physiol. (Lond.) 113:157-168.

Hunt JN, Stubbs DF 1975. The volume and energy content of meals as determinants of gastric emptying. J. Physiol. 245: 209-25. lmaizumi M, Hama K 1969. An electronmicroscopic study on lhe interstitial cells of the gizzard in the love-bird (Urolocha domestica). Z. Zelllorsch 97: 351-357. lnui A, Oya M, Okita M, lnoue T, Sakatani N, Morioka H, Shii K, Yokono K 1988. Peptide YY receptors in the brain. Biochem. Biophys. Res. Comm. 15O: 25-32. lrwin DA 1931. The anatomy of Auerbach's plexus. Am. J. Anatomy 49:141-166. lsenberg Jl, Csendes A 1972. Effecls of octapeptide of cholecystokinin on canine pylorus pressurê. Am. J. Physiol. 222: 428-431.

Jain NK, Boivin M, Zinsmeister AR, Brown ML, Malagelada J-R, DiMagno EP 1989. Effect of ileal perfusion of carbohydrate and amylase inhibitor on gastrointestinal hormones and emptying. Gastroenterology 96: 377-387. 175

Jian R, Vigneron V, Najean Y, Bernier J 1982. Gastric emptying and intragastric distribution of lipids in man. Dig. Dis. Sci. 27:705-711.

Jaup BH, Dotevall G 1981. The effect of pirenzepine and L-hyoscyamine on gastric emptying and salivary secretion in healthy volunteers. Scand. J. Gastroenterol. 16:. 769-773.

Johnson AG 1975. The pylorus. ln: Thomas, Mann Eds., Alimentary sphincters and their disorders. McMillan Publishers Ltd., London, pp. 101-133.

Johnson AG, Eyre-Brook lA 1984. New developments in the evaluation of gastroduodenal motility with special reference to duodenogastric reflux and its clinical significance. Scand. J. Gastroenterol. 19 (Suppl. 96): 27-35.

Johnson AG, Linhardt GE, Eyre-Brook lA, Smallwood RH, Labo G, Bortolotti M Eds. 1983 Gastrointestinal Motility. Cartinia lnternational, Verona, pp. 251-254.

Kaye MD, Mehta SJ, Showalter JP 1976. Manometric studies of the human pylorus. Gastroenterology 70: 477-48O.

Kaye MD, Keinke O, Ehrlein H-J 1983. Effect of oleic acid on canine gastroduodenal motility, pyloric diameter and gastric emptying. Quart. J. Exp. Physiol. 68: 675-686.

Keet AD 1957. The prepyloric contractions in the normal stomach. Acta. Radiol.48:'413-424.

Keinke O, Ehrlein H-J 1983. Effect of oleic acid on canine gastroduodenal motility, pyloric diameter and gaslric emptying. Quart. J. Exp. Physiol. 68: 675-686.

Keinke O, Schemann M, Ehrlein H-J 1984. Mechanical factors regulating gastric emptying of viscous nutrient meals in dogs. Quart. J. Exp. Physiol. 69: 781-795.

Keinke O, Wulschke S, Ehrlein H-J 1986. Neurotensin slows gastric emplying by a transient inhibition of gastric and a prolonged inhibition of duodenal motility. Digestion 34:- 281-288.

Kellow JE, Gill RC, Wingate DL 1987. Modulation of human upper gastrointestinal motility by rectal distension. Gut 28: 864-868 176

Kelly KA 1980. Gastric emptying of liquids and solids: roles of proximal and distal stomach. Am. J. Physiol. 234: G71-G76.

Kelly KA, Code CF 1971. Canine gastric pacemaker. Am. J. Physiol. 22O: 112-118

Kelly KA 1974. Dfiferential responses of the canine gastric corpus and antrum to electrical stimulation. Am. J. Physiol. 226: 23o-234.

Kelly KA, Code CF, Elveback LR 1969. Patterns of canine gastric electric activity. Am. J. Physiol. 217: 461-47O.

Khamis B, Kennedy C, Finucane J, Doyle JS 1978. Transmucosal potential difference: diagnoslic value in gastrooesophageal reflux. Gut 19: 396-398.

King PM, Adam RD, Pryde A, McDicken WN, Heading RC 1984. Relationships of human antroduodenal motility and transpyloric fluid movement: non-invasive observations with real- time ultrasound. Gut 25: 1384-1391.

King PM, Heading RC, Pryde A 1985. Coordinated motor activity of the human gastroduodenal region. Dig. Dis. Sci. 30: 219-224.

King PM, Pryde A, Heading RC 1987. Transpyloric fluid movement and antroduodenal motility in patients with gastro-oesophageal reflux. Gut 28: 545-548.

Kleibeuker JH, Beekhuis H, Jansen JBMJ, Piers DA, Lamers CBHW 1988a. Cholecystokinin is a physiological mediator of fat-induced gastric emptying in man. Eur. J. Clin. lnvest. 18: 173-177.

Kleibeuker JH, Beekhuis H, Piers DA, Schaffalitzky de Muckadell OB 1988b. Retardation of gastric emptying of solid food by secretin. Gastroenterology 94: 122-126.

Kobayashi M, Nagai T, Prosser CL 1966. Electrical interaction between muscle layers of cat intestine. Am. J. Physiol. 211: 1281-1291.

Kondo Y, lhara O, Yamamoto O 1985. Effects of various surgical procedures on gastric emptying of glucose solutions. Jap. J. Smooth Muscle Res.21 (Suppl. 1):85-91. 177

Konturek SJ, Bilski J, Thor P, Czarnobilski K, Szoke B, Schally AV 1985. Gastrointestinal secrstory, motor and circulatory effects of corticotropin releasing factor (CRF). Life Sci. 37: 1231 -1240.

Kumar D, Ritman EL, Malagelada J-R 1987. Three-dimensional imaging of the stomach: role of the pylorus in the emplying of liquids. Am. J. Physiol. 253 (Gastrointest. Liver Physiol. 16): G79-G85.

Kurosawa M, Malsuo Y, Ohno H, Sato A 1988. Vagal nerve contribution to the calcitonin induced changes in intrapyloric pressure in anaesthetised rats. Newssci. Lett.86:322-327.

Langley JN 1898. On inhibitory fibres in the vagus for the end of the oesophagus and the stomach. J. Physiol. 23:. 4O7-414.

Lavigne ME, Wiley ZD, Meyer JH, Martin P, MacGregor lL 1978. Gaslric emptying rates of solid food in relation lo body size. Gastroenterology 74: 1258-1260.

Leek BF 1972. Abdominal visceral receptors. Chapter 4ln: Neil E Ed., Handbook of Sensory Physiology, Volume 3/1, Enteroceptors. Springer-Verlag, Berlin, pp.113-160.

Leidberg G, Nielson KC, Owman CH, Sjoberg N-O 1973. Adrenergic contribution to the abdominal vagus nerves in the cat. Scand. J. Gastroenterol. 8: 177-180.

Lerman SH, Mason GR, Bathon EM, Ormsbee HS 1981. Pyloric molor response to sympathetic nerve stimulation in dogs. Surgery 89: 460-465.

Lidberg P, Dahlstrom A, Ahlman H 1984. ls S-HT a mediator in lhe motor control of the feline pylorus? Scand. J. Gastroenterol. 19(3): 321-328.

Liddle RA, Goldfine lD, Rosen MS, Taplitz RA, Williams JA 1985. Cholecystokinin bioactivity in human plasma. Molecular forms, responses to feeding and relationship to gallbladder contraction. J. Clin. lnvest. 75: 1144-1152.

Liddle RA, Morita ET, Conrad CK, Williams JA 1986. Regulalion of gastric emptying in humans by cholecystokinin. J. Clin. lnvest. 77: 992-996. 178

Lin HC, Doty JE, Reedy TJ, Meyer JH 1988. lnhibition of gastric emptying by acids depends on the length of gut exposed lo lhe acid. Gaslroenlerology 94:. 4262 (abslract).

Lin HC, Doty JE, Reedy TJ, Meyer JH 1989a. Inhibition of gastric emptying by glucose depends on the length of gut exposed to the nutrient. Am. J. Physiol. 256 (Gastrointest. Liver. Physiol.

19): G404-G41 1.

Lin HC, Doty JE, Reedy TJ, Meyer JH 1989b. lnhibition of gastric emptying by fatty acid depends on length of gut exposed to the nutrient. Gastroenterology; 96: A304 (abstract).

Linhardt GE, Robinson P, Smallwood RH, Johnson AG 1981. Does the human pylorus close in response to isolated duodenal cap contraction? Gut 22: 4882 (abstract).

Lipshutz W, Cohen S 1972. lnteraction of gastrin I and secretin on gastrointestinal circular muscle. Am. J. Physiol. 222: 775-781.

Longstreth GF, Malagelada J-R, Go VLW 1975. The gastric responss to a transpyloric duodenal tube. Gut '16: 777-78O.

Loo FD, Palmer DW, Soergel KH, Kalbfleisch JH, Wood CM 1984. Gastric emptying in patients with diabetic mellitus. Gastroenterology 86: 485-494.

Looijen R, Feenstra BWA, van der Schee EJ, Roelofsen HJ, Bruining HA 1987. Postoperative motility patterns in stomach and duodenum. Oxford Motility Meeting, p. 130.

Louckes HS, Quigley JP, Kersey J 1960. lnductograph method for recording muscle activity, especially pyloric sphincter physiology. Am. J. Physiol. 199: 301-310.

Lovallo W 1985. The cold pressor test and autonomic function: a review and integration Psychophysiology 12: 268-282.

Lucey MR, Wass JAH, Fairclough P, Webb J, Medback S, Rees LH 1985. Autonomic regulation of postprandial plasma somatoslatin, gastrin and insulin. Gut 26: 683-688.

Ludlke FE, Golenhofen K, Becker HD 1983. Mechanical activity of isolated human pyloric muscle, in comparison with canine pylorus. ln: Roman C Ed., Gastrointestinal Motility. pp. 87-94. 179

Ludwick JR, Wiley JN, Bass P 1969. Pyloroplasly and vagotomy. Early effects on antral and duodenal contractile activity. Arch. Surg. 99: 553-559.

Lundberg JM, et al 1979. Substance P, vasoactive intestinal polypeptide, and enkephalin-like immunoreactivity in the human vagus nsrve. Gaslroenterology 77: 468-471.

McOoy EJ, Bass P 1963. Chronic electrical activity of gastroduodenal area: effect of food and certain catecholamines. Am. J. Physiol. 2O5:- 439-445.

McOullough A, Miller L, Service F, Go V 1983. Effect of graded intraduodenal glucose infusions on the releass and physiological action of gastric inhibitory polypeptide. J. Clin. Endocrinol. Metab. 56: 234-241.

MacGregor l, Parenl J, Meyer JH 1977. Gastric emptying of liquid meals and pancreatic and biliary secretions after subtotal gastrectomy or truncal vagotomy and pyloroplasty in man.

Gastroenterology 72: 1 95-205.

McHugh PR, Moran TH 1986. The stomach, cholecystokinin and satiety. Fed. Proc.45: 1384-

1 390.

McLure CW, Reynolds L, Schwartz CO 1920. On the behaviour of the pyloric sphincter in normal man. Arch. lnt. Med. 29: 1.

McRae S, Younger K, Thompson DG, Wingate DL 1982. Sustained mental slress alters human jejunal motor activity. Gut 23: 4O4-4O9.

McShane AJ, O'Morain C, Lennon JR, Wakley JB, Alton BG 1980. Atraumatic, non-distorting pyloric sphincter studies. Gut 21: 826-828.

Maddern GJ, Horowitz M, Jamieson GG, Chatterton BE, Collins PJ, Roberts-Thompson P 1984. Abnormalities of oesophageal and gastric emptying in progressive systemic sclerosis. Gastroenterology 87: 922-526.

Mangnall YF 1989. Clinical uses of applied potential tomography and impedance epigastrography as methods of measuring gastric emptying. ln: Read NW Ed., Gastrointestinal Motility. Which Test? Wrightson Biomedical Publishing Ltd., Petersfield, pp. 91-98. 180

Malagelada JR, Go VLW, Summerskill WHJ 1979. Different gastric, pancreatic and biliary r€sponses to solid-liquid or homogenized meals. Am. J. Dig. Dis.24: 101-110.

Malagelada J-R, Rees WDW, Mazzotta LJ, Go VLW 1980. Gastric motor abnormalities in diabetic and post-vagotomy gastroparesis: effect of metoclopramide and bethanecol. Gastroenterology 78: 286-293.

Malmud LS, Fisher RS, Knight LC, Rock E 1982. Scintigraphic evaluation of gastric emptying. Semin. Nucl. Med. 12: 116-125.

Mamtora H, Thompson DG 1989. Gastric Ullrasound. ln: Read NW Ed., Gastrointestinal Motility Which Test? Wrightson Biomedical Publishing Ltd., Petersfield, pp. 99-104.

Mearin F, Azpiroz F, Malagelada J-R 1986. Measurement of resistance to flow across antroduodenal area during fasting. Am. J. Physiol.250 (Gastrointest. Liver Physiol. 13): G773- G780.

Mearin F, Azpiroz F, Malagelada J-R 1987. Pyloric contribution to antroduodenal resistance to flow in the conscious dog. Am. J. Physiol. 253 (Gastrointest. Liver Physiol. 16): G72-G78.

Mearin F, Cucala M, Azpiroz F, Malagelada J-R 1989. Origin of gastric symptoms in functional dyspepsia. Gastroenterology 97: A337.

Meeroff JC, Go VLW, Phillips SF 1975. Control of gastric emptying by osmolarity of duodenal contents in man. Gastroenterology 68: 1144-1151.

Mei N 1970. Mechanorecepteurs vagaux digestifs chez le chat. Experimental Brain Research 11 s02-51 4.

Mei N 1978. Vagal glucoreceptors in the small intestine of the cat. J. Physiol. 282:485-506.

Meschan l, Quigley JP 1938. Spontaneous motility of the pyloric sphincter and adjacent regions of the gut in the unanesthetized dog. Am. J. Physiol. 21 1: 350-357.

Meyer BM, Bellinger C, Jansen JMBJ, Rovati LC, Werth BA, Hildebrand P, Zach D, Stalber GA

1989. Role of cholecystokinin in regulation of gastrointestinal motor function. Lancet July 1: 12- 15. 181

Meyer JH 1987. Motility of the stomach and gastroduodenal junction. ln: Johnson LR Ed., Physiology of the Gastrointestinal Tract,2nd Edition. Raven Press, New York, pp. 613-629.

ee'Tc-tagged Meyer JH, MacGregor lL, Guellar R, Martin P, Cavalieri 1976. chicken liver as a marker of solid food in the human stomach. Dig. Dis. 21:. 296-304.

Meyer JH, Thomson JB, Cohen MB, Schadchehr A, Mandiola SA 1979. Sieving of solid food by the canine stomach and sieving after gastric surgery. Gaslroenterology 76: 804-813.

Meyer JH, Dressman J, Fink AS, Amidon G 1985. Effect of size and density on gastric emplying of indigestible solids. Gastroenterology 89: 805-813.

Miller J, Kauffman G, Elashoff J, Ohashi H, Carter D, Meyer JH 1981. Search for resistances controlling canine gastric emptying of liquid meals. Am. J. Physiol. 241 (Gaslrointest. Liver Physiol. 4): G403-G415.

Milton GW, Smith AWM, Armslrong HIO 1955. The origin of rhythmic electropotential changes in the duodenum. Quart. J. Exp. Physiol. 40:79.

Mir SS, Mason GR, Ormsbee HS 1977. An inhibitory innervalion at the gastroduodenal junction in anesthetized dogs. Gastroenterology 73:. 432-434.

Mir SS, Mason GR, Ormsbee HS 1978. Vagal influence on duodenal motor activity. Am. J. Surg. 135:97-101.

Mir SS, Telford GL, Mason GR, Ormsbee HS 1979. Noncholinergic nonadrenergic inhibitory innervation of the canine pylorus. Gastroenterology 76: 1443-1448.

Moore JG, Dubois A, Christian PE, Elgin D, Alazraki N 1986. Evidence for a mid-gastric transverse band in humans. Gastroenterology 91: 540-545.

Moore JG, Datz FL, Christian PE, Greenberg E, Alazraki N 1988. Effect of body posture on radionuclide measurements of gastric emptying. Dig. Dis. Sci. 33: 1592-1595.

Moran TH, McHugh PR 1982. Cholecystokinin suppresses food intake by inhibiting gastric emptying. Am. J. Physiol. 242: R491-R497. 182

Morgan KG, Schmallz PF, Szurszewski JH 1978. The inhibitory effects of vasoactive intestinal polypeptide on the mechanical and electrical act¡vity of canine antral smooth muscle. J. Physiol. (Lond.) 282: 437-45o.

Morgan KG, Szurszewski JH 1980. Mechanisms of phasic and tonic actions of pentagastrin on canine gastric smooth muscls. J. Physiol. (Lond.) 3O1:. 229-242.

Muller-Lissner SA, Blum A 1984. To-and-fro movements across the canine pylorus. Scand. J Gastroenterol. 19 (Suppl. 92): 1-3.

Muller-Lissner SA, Schindlebeck NE 1984. Relevance of gastroduodenal coordination in duodenogastric reflux. ln: Van Nueten, Schuurkes, Akkermans Ed., Gastroduodenal Coordination.

Muller-Lissner SA, Fimmel CJ, Will N, Muller-Duysing W, Heinzel F, Blum AL 1982. Effect of gastric and transpyloric tubes on gastric emptying and duodenogastric reflux. Gastroenterology 83: 1276-1279.

Munk JF, Gannaway RM, Hoare M, Johnson AG 1978. Direct measurement of pyloric diameter and tone in man and their response to cholecystokinin. ln: Duthie HL Ed., Gastrointestinal Motility in Health and Disease. MTP Press, Lancaster, pp. 349-356.

Neal DE, Williams NS, Barker MCJ, King RFGJ 1984. The effect of resection of the distal ileum on gastric emptying, small bowel transit and absorption after proctocolectomy. Br. J. Surg. 71: 666-670.

Nimmo WS 1976. Drugs, diseases and altered gastric emptying. Clin. Pharmacokinetics 1:189- 203.

O'Brien JD, Thompson DG, Burnham WR, Holly J, Walker E 1987. Action of centrally mediated autonomic stimulation on human upper gastrointestinal transit: a comparative study of two stimuli. Gut 28: 960-969.

Ok¡ M, Daniel EE'1974. Ultrastructural basis for electrical coupling in the dog stomach. ln: Daniel EE Ed., Proceedings of the 4th lnternational Symposium on Gastrointestinal Motility. Mitchell Press Ltd., Vancouver, pp. 85-95. 183

Orlando RC, Powell DW, Kinnard HB, Carney CN, Jones JD, Bozymski EM 1982. Esophageal

"i potential difference measurements in esophageal disease. Gastroenterology 83: 1926-1032.

Ormsbee HS, Bass P 1976. Gastroduodenal molor gradients in the dog after pyloroplasty. Am. J Physiol. 230: 389-397.

Pagani FD, Norman WP, Kasbeker DK, Gillis RA 1985. Localization of siles within dorsal motor nucleus of vagus that affect gastric emptying. Am. J. Physiol. 249 (Gastrointest. Liver Physiol. 12): G63-G84.

Paintal AS 1954. A study of gastric stretch receptors. Their role in the peripheral mechanism of satiation of hunger and thirst. J. Physiol. (Lond.) 126:. 255-270.

Pandolfo N, Bortolotti M, Nebiacolombo C, Labo G, Mattioli F 1979. Prolonged manometric study of the gastroduodenal junction in man. Digestion 19: 86-92.

Papasova MP, Nagai T, Prosser CL 1968. Two-component slow waves in smooth muscle of cat stomach. Am. J. Physiol. 214: 695-702.

{ tl ii Papasova M, Kortesova N, Atanassova E 1982. Parlicipalion of the pyloric sphincter in the coordination of the electrical activity in the stomach and duodenum. ln: Weinbeck M Ed., Motility of the Digestive Tract. Raven Press, New York, pp. 201-206.

Papasova M 1989. Sphincteric function. ln: Wood JD Ed., Handbook of Physiology, Section 6, The Gastrointestinal Tract, Chapter 26. American Physiological Society, Bethesda, Maryland, pp. 987 -1024.

Pappas TN, Debas HT, Chang AM, Taylor lL 1986. Peptide YY release by fatty acids is sufficient to inhibit gastric emptying in dogs. Gastroenlerology 91: 1386-1389.

Parr NJ, Grime JS, Baxter JN, Critchley M, Mackie CR 1987. Small bowel resistances and the duodenal brake. Gut 28: 950-954.

Pascaud XB, Genton MJH, Bass P 1978. Gastroduodenal contractile activity in fed and fasted unrestrained rats. ln: Duthie HL Ed., Gastrointestinal Motility in Health and Disease. Lancaster, MTP Press, pp. 637-645. 184

Phaosawasdi K, Fisher RS 1982. Hormonal effects on the pylorus. Am. J. Physiol.243: G330- G33s.

Plotkin A, Reynolds JC 1989. Extrinsic peptidergic innervation of the rat pyloric sphincter Gastroenterology 96: 4690 (abstract).

Polak J, Colling l, Bloom S, Pearse A 1971. lmmunofluorescent localization of secretin and enteroglucagon in human intestinal mucosa. Scand. J. Gastroenterol. 6: 739-744.

Pope CE 1967. A dynamic test of sphincter strength; its application to the lower esophageal sphincter. Gastroenterology 52: 779-786.

Prove J, Ehrlein H-J 1982. Motor function of gastric antrum and pylorus for evacuation of low and high viscosity meals in dogs. Gut 23: 150-156.

Quigley JP, Werle JM 1940. The correlation of antral and bulbar pressures with fluoroscopic observations during gastric evacuation. Am. J. Dig. Dis. 10: 434-435.

]t i1 Quigley JP, Read MR, Radzow KH, Meschan l, Werle JM 1942. The effect of hydrochloric acid on ¡,lf the pyloric sphincter, the adjacent portions of the digestive tract and on the procêss of gastric evacuation. Am. J. Physiol. 137: 153-159.

Raybould HE 1989. Cholecystokinin and capsaicin-sensitive vagal afferent fibers mediate changes in gastric motility following intraduodenal infusion of protein in the rat. Gastroenterology 96: 4410 (abstract).

Raybould HE, Dent J, Reddy SN, Tache Y 1989a. Central and peripheral peptidergic regulation of rat pyloric motor function. Gastroenterology 96: A410 (abstract).

Raybould HE, Kolve E, Tache Y 1989b. Central nervous system action of calcitonin gene-related peptide (CGRP) to inhibit gastric motor function by adrenergic mechanisms in the ral. Gaslroenterology 96: A410 (abstract).

Read NW 1989. Tests of gastric motility - summary and conclusions. ln: Read NW Ed., Gastrointestinal Motility. Which Tesl? Wrightson Biomedical Publishing Ltd., Petersfield, pp. i 1 33-1 40.

Ì 185

Read NW, Fordtran JS 1979. The role of intraluminal junction potentials in the generation of the gastric potential difference in man. Gaslroenterology 76: 932-938.

Read NW, Al Janabi MN, Bates TE, Barber DC 1983. Effect of gastrointestinal intubation on lhe passage of a solid meal through the stomach and small intestine in humans. Gastroenterology 84:

1 s68-1 572.

Read NW, McFarlane A, Kinsman Rl, Bates TE, Blackhall NW, Farrar GBJ, Hall JC, Moss G, Morris AP, O'Neill B, Welch l, Lee Y, Bloom SR 1984. Effect of infusion of nutrient solutions into the ileum on gastrointestinal transit and plasma levels of neurotensin and enteroglucagon. Gastroenterology 86; 274-80.

Reddy SN, Daniel EE 1985. Alpha-adrenergic control of duodenal inhibition by antral contractions. Dig. Dis. Sci. 30: 789 (abstract).

Reynolds JC, Ouyang A, Cohen S 1984. Evidence for an opiate-mediated pyloric sphincter reflex Am. J. Physiol. 246 (Gastrointesl. Liver Physiol. 9): G130-G136.

-t ù( Reynolds JC, Ouyang A, Cohen S 1985. Opiate nerves mediate feline pyloric response to intraduodenal amino acids. Am. J. Physiol.248 (Gastrointest. Liver Physiol. 11): G307-G312. 1

Riddle P, Treacy J, Jamieson GG, Dent J 1989. The effect of truncal vagolomy on pyloric motility and liquid emptying in the pig. Gastroenterology (Submitted).

Roman C 1982. Nervous control of esophageal and gastric motility. ln: Bertaccini G Ed., '1' Mediators and Drugs in Gastrointestinal Motility 1. pp. 23-278.

Rouillon JM, Azpiroz F, Malagelada J-R 1989. Responsiveness of the jejunum to intestino- intestinal reflexes in humans. Gastroenterology 96: A427 (abstract).

Rovelstad RA 1976. The incompetent pyloric sphincter. Bile and mucosal ulceration. Am. J. Dig Dis. 21 : 165-173.

Ruckebusch Y, Malbert CH 1986. Physiological characteristics of ovine pyloric sphincter. Am. J I Physiol. 251 (Gastrointest. Liver Physiol. 14): c804-c810.

* 186

Sarna SK 1975. Gastrointeslinal electrical aclivity: terminology. Gastroenterology 68: 1631-

1 63s.

Sarna SK 1989. ln vivo myoelectrical activity: methods, analysis, and interpretation. ln: Schultz, Wood, Raurer Eds., Handbook of Physiology, Seclion 6, Volume 1, The Gastrointestinal System. American Physiological Sociely, Bethesda, Maryland, pp. 817-864.

Sarna SK, Daniel EE, Kingma YJ 1572. Simulation of the electric-control activity of the stomach by an array of relaxalion oscillators. Am. J. Dig. Dis. 17:299-310.

Sarna SK, Condon RE 1984. Morphine-initíaled migrating myoelectrical complexes in the fed state in dogs. Gastroenterology 86: 662-669.

Savage AP, Adrian TE, Carolan G, Chatterjee VK, Bloom SR 1987. Effects of peptide YY (PYY) on mouth to caecum intestinal transit time and on the rate of gastric emptying in heallhy volunleers. Gut 28: 166-170.

Schaffalitzky de Muckadell O, Fahrenkrug J 1978. Secretion pattern of secretin in man regulation by gastric acid. Gut 19: 812-818. 'T

Scheurer V, Varga L, Drack E, Burki H-R, Halter F 1983. Measurement of cholecystokinin oclapeptide-induced motility of rat antrum, pylorus, and duodenum in vitro. Am. J. Physiol. 244 (Gastrointest. Liver Physiol. 7): G261-G265.

Schulze-Delrieu K, Brown CK 1985. Emptying of saline meals by the cat stomach as a function of pyloric resistance. Am. J. Physiol. 249 (Gastrointest. Liver Physiol. 12): G725-G732.

Schulze-Delrieu K, Shirazi SS 1983. Neuromuscular differentiation of the human pylorus Gastroenterology 84: 287-292.

Schulze-Delrieu K, Wall JP 1983. Determinants of flow across isolated gastroduodenal junctions of cats and rabbits. Am. J. Physiol. 245: G257-G264.

Schulze-Delrieu K, Ehrlein HJ, Blum AL 1984. Mechanics of lhe pylorus. ln: Akkermans LMA, Johnson AG, Read NW Eds., Gastric and Gastroduodenal Motility. Praeger Publishers, pp. 87- i 102.

þ 187

Schuurkes JAJ, van Neuten JM 1984. Gastroduodenal coordination. ln: Akkermans LMA, Johnson AG, Read NW Eds., Gastric and Gastroduodenal Motility. Praeger Publishers, pp 103-127.

Seide K, Ritman EL 1984. Three-dimensional dynamic x-ray-computed lomography imaging of stomach motility. Am. J. Physiol. 247 (Gaslrointest. Liver Physiol. 10) G574-G581.

Shay H, Gershon-Cohen J 1934. A study of pyloric control: roles of acid and alkali. Surg. Gyn Obstet. 58: 935-955.

Sheiner HJ, Quinlan MF, Thompson lJ 1980. Gastric motility and emptying in normal and post- vagotomy patients. Gut 21: 753-759.

Skandalakis LJ, Skandalakis JE 1986. The history and surgical anatomy of the vagus nerve Surg. Gyn. & Obst. 162: 75-85.

Smith AWM, Code CF, Schlegel JF 1957. Simultaneous cineradiographic and kymographic studies of human gastric antral motility. J. Appl. Physiol. 11: 12-16.

Smout AJPM, Van der Schee EJ, Grashuis JL 1980. What is measured in electrogastrography? Dig. Dis. Sci. 25: 179-187.

Spielberger CD 1983. Manual for the state-trait anxiety inventory (form Y). Consulting Psychologists Press lnc., Palo Alto, California

Spiller RC, Trotman lF, Higgins BE 1984. The ileal brake - inhibition of jejunal motility after ileal fat perfusion in man. Gut 25: 365-374.

Spiller RC, Lee YC, Ralphs DNL, Stewart JS, Bloom SR, Silk DBA 1987. Delayed moulh-caecum transit of a lactulose labelled liquid test meal in patients with steaorrhoea caused by partially treated coeliac disease. Gut 28: 1275-82.

Spiller RC, Trotman lJ, Adrian TE, Bloom SR, Misiewicz JJ, Silk DBA 1988. Further characterisation of the "ileal brake' reflex in man - effect of ileal infusion of partial digests of fat, prolein, and starch on jejunal motility and release of neurotensin, enleroglucagon, and peptide YY. Gut 29: 1042-1051. I

h 188

Stacher G, Bergmann H, Harlik E, Schmierer G, Schneider C 1984. Effects of oral cyclotropium bromide, hyoscine N-butylbromide and placebo on antral motor activity in healthy men. Gut 25: 485-490.

Stanghellini V, Malagelada J-R, Zinsmeister AR, Go VLW, Kao PC 1983. Stress-induced gastroduodenal motor disturbances in humans: possible humoral mechanisms. Gastroenterology 85: 83-91.

Stanghellini V, Malagelada J-R, Zinsmeister AR, Go VLW, Kao PC 1984. Effect of opiate and adrenergic blockers on the gut motor response to centrally acting stimuli. Gastroenterology 87: 1104-1113.

Stemper TJ, Cooke AR 1976. Effecl of a fixed pyloric opening on gastric emptying in the cat and dog. Am. J. Physiol. 230: 813-817.

Stern RM, Koch KL, Stewart WR, Lindblad lM 1987. Spectral analysis of tachygaslria recorded during motion sickness. Gaslroenterology 92: 92-97.

Subissi A, Guelfi M, Maggi CA, Meli A 1985. lnfluence of "selective" antimuscarinic agents on gastric acid secretion and motility. Arch. lnt. Pharmacodyn. 274:313-319.

Szurszewski JH 1987. Electrical basis for gastrointestinal motility. ln: Johnson LR Ed., Physiology of the Gastrointestinal Tract, Second Edition. Raven Press, New York, pp. 383-422.

Takayama K, lshikawa N, Muira M 1982. Sites of origin and lermination of gastric preganglionic neurones: a HRP study in the rat. J. Aut. Nerv. Syst. 6: 211-223.

Tatemoto K, Nakano l, Makk G, Angwin P, Mann M, Schilling J, Go VLW 1988. lsolation and primary structure of human peptide YY. Biochem. Biophys. Res. Comm.157:713-717.

Taylor I 1985. Distribution and release of peptide YY in dog measured by specific radio- immunoassay. Gastroenterology 88: 731-737.

Telford GL, Mir SS, Mason GR, Ormsbee HS 1979. Neural control of the canine pylorus. Am. J. Surg 137: 92-97.

Thomas JE 1957. Mechanics and regulation of gastric emptying. Physiol. Review 37: 453-474. 189

Thomas JE, Crider JO 1935. Rhythmic changes in duodenal motility associated wilh gastric peristalsis. Am. J. Physiol. 11 1 : 124-129.

Thomas JE, Crider JO, Mogan CJ 1934. A study of reflexes involving the pyloric sphincter and antrum and their role in gastric evacuation. Am. J. Physiol 108: 638-700.

Thomas JE, Kuntz A 1926. A study of gastro-intestinal motility in relation to the enteric nervous system. Am. J. Physiol. 76: 606-626.

Thompson DG, Richelson E, Malagelada J-R 1982. Perturbation of gastric emptying and duodenal motility through the central nervous system. Gastroenterology 83: 1200-1206.

Thompson DG, Richelson E, Malagelada J-R 1983. Perturbation of gastrointestinal function by cold stress. Gul 24: 277-283.

Thompson DG, Wingate DL 1988. Effect of osmoreceptor stimulation on human duodenal motor activity. Gut 29: 173-180.

Thuneberg L 1982. lnterstitial cells of Cajal. lntestinal pacemaker cells. Adv. Anal. Embryol. Cell Biol. 71 : 1-130.

Thuneburg L 1989. lnterstitial cells of Cajal. ln: Wood JD Ed., Handbook of Physiology, Section 6 The Gaslrointestinal Tract, Chapter 10. American Physiological Society, Bethesda, Maryland, pp. 349-386

Torgersen J 1942. The muscular build and movements of the stomach and duodenal bulb. Acta Radiologica Supplementum 45. Fabritius & Sonners Boktrykken, Oslo, pp. 1-191.

Tothill P, Mcloughlin GP, Heading RC 1978. Techniques and errors in scintigraphic measurements of gastric emptying. J. Nucl. Med. 19: 256-261.

Tougas G, Anvari M, Richards AD, Dent J, Somers S, Stevenson GW 1987. Relationship of pyloric motility to transpyloric flow in healthy subjects. Gastroenterology 92: 1673 (abstract).

Tougas G, Allescher HD, Li YY, Daniel EE, DentJ, Hunt RH 1988a. Effect of duodenal acidification on pyloroduodenal motor activity in humans. Gaslroenterology 94: A464 (abstract). 190

Tougas G, Bovell KT, Collins SM, Howard JA, Dent J, Hunt RH 1988b. Relationship of circulating cholecystokinin (CCK) and pyloric motor response after intraduodenal triglyceride infusion in humans. Gastroenterology 94: 4464 (abstract).

Treacy PJ, Jamieson GG, Dent J 1988a. The effect of pylorus excision on gastric emptying of a digestible solid in the pig. J. Gastroenterol. Hepatol. 3 (Suppl.1):21 (abstract).

Treacy PJ, Jamieson GG, Dent J 1988b. Duodenal intramural nerves are major controllers of pyloric motility and gastric emptying. Gastroenterology 94: 4464 (abstract).

Valenzuela JE, Defilippi C 1976. Pyloric-sphincter studies in peptic-ulcer patienls. Pylorus in peptic ulcer. Dig. Dis. 21: 229-232.

Valenzuela JE, Defilippi C 1981. lnhibition of gastric emptying by secretin, the octapeptide of cholecystokinin and intraduodenal fat. Gastroenterology 81: 898-902.

Valenzuela JE, Defilippi C, Csendes A 1976. Manometric studies on the human pyloric sphincter. Effect of smoking, metoclopramide, and atropine. Gastroenterology 70: 481-483.

Valori RM, Collins SM, Daniel EE, Reddy SN, Shannon S, Jury J 1986. Comparison of methodologies for lhe measurement of antroduodenal activity in the dog. Gastroenterology 91: 546-553.

Van der Schee EJ 1974. Electrogastrography. Signal analytical aspects and interpretation Doctoral Thesis.

Van der Schee EJ, de Bakker JV, Zwamborn AW 1981. Transducer for in vivo measurement of the inner diameter of arteries in laboratory animals. Med. & Biol. Eng. & Comput. 19: 218-222.

Van der Schee EJ, Geldof H 1985. A new sensor for the measurment of inlerdigestive and digestive gastric motility in man. Dig. Dis. Sci. 30: 799 (abstract).

VanDeventer G, Thomson J, Graham LS, Thomasson D, Meyer JH 1983. Validation of correction for errors in collimation during measurement of gastric emptying of nuclide-labeled meals. J. Nucl. Med. 24: 187-196. 191

Walsh JD 1987. Gastrointestinal hormones. ln: Johnson LR Ed., Physiology of the Gastro- intestinal Tract, 2nd Ed. Raven Press, New York, pp. 181-253.

Wattchow DA, Cas DT, Furness JB, Costa M, O'Brien PE, Little KE, Pitkin J 1987. Abnormalities of peptide-containing nerve fibres in hypertrophic pyloric stenosis. Gastroenterology 92: 438- 444.

Wattchow DA, Furness JB, Costa M 1988. Distribution and coexislence of peptides in nerve fibres of the external muscle of the gastrointestinal tract. Gastroenterology 95: 32-41.

Weber J, Kohatsu S 1970. Pacemaker localisation and electrical conduction patterns in the canine stomach. Gastroenterology 59: 717-726.

Weisbrodt NW, Wiley JN, Overholt BF, Bass P 1969. A relation between gastroduodenal muscle contractions and gastric emptying. Gut 10: 543-548.

Welch lMcL, Cunningham KM, Read NW 1988. Regulation of gastric emptying by ileal nutrients in humans. Gastroenterology 94: 4O1-4O4.

Werth B, Meyer JH, Beglinger C, Hildebrand P, Stalder AGA 1989. Loxiglumide (Lox), a new cholecystokinin (CCK) receptor antagonist, affects regional gastrointestinal transit times in healthy volunleers. Gastroenterology 96: A542 (abstract).

Wheelon H, Thomas JE 1921. Rhythmicity of the pyloric sphincter. Am. J. Physiol. 54:. 460-473.

Wheelon H, Thomas JE 1922. Observations on the motility of the duodenum and the relation of duodenal activity lo that of lhe pars pylorica. Am. J. Physiol. 59: 72-96.

White CM, Poxon V, Alexander-Williams J 1981. A study of motility of normal human gastroduodenal region. Dig. Dis. Sci. 26: 609-617.

White CM, Poxon V, Alexander-Williams J 1983. Effects of nutrient liquids on human gastroduodenal motor activity. Gut 24: 1109-1116.

Wilbur BG, Kelly KA 1973. Effect of proximal gastric, complete gastric and truncal vagotomy on canins electric activity, motility and emptying. Ann. Surg. 178: 295-303. 192

Wiley J, Tatum D, Keinath R, Owyang C 1988. Participation of gastric mechanoreceptors and intestinal chemoreceptors in the gastrocolonic response. Gastroenterology 94: 1144-1149.

Wilkinson AR, Johnson D 1973. Effects of truncal, selective and highly selective vagotomy on gastric empty¡ng of a food-barium meal in man. Ann. Surg. '178:190-193.

Williams I 1962. Closure of the pylorus. Br. J. Radiol. 35: 653-670

Williams PL, Warwick R 1980. The Stomach. ln: Gray's Anatomy, Chapter 8 (Splanchnology). pp.

1 333-1 342.

Yamagishi T, Debas HT 1978. Cholecystokinin inhibils gastric emptying by acting on both proximal stomach and pylorus. Am. J. Physiol. 234 (Endocrinol. Metab. Gastrointest. Physiol. 3): E37s-E378.

Yammamolo T, Satomi H, Zke H, Takahashi K 1977. Evidence of the dual innervation of the cat stomach by the vagal dorsal motor and medial solitary nuclei as demonstrated by the horseradish peroxidase method. Brain Res. 122: 125-131.

You CH, Chey WY, Lee KY, Menguy R, Bortoff A 1981. Gastric and small intestinal myoelectric dysrhythmia associated with chronic intractable nausea and vomiting. Ann. lnt. Med. 95: 449- 451.

You CH, WY Chey 1984. Study of electromechanical activity of the stomach in humans and in dogs with particular attention to tachygastria. Gastroenterology 86: 1460-1468.

Youle MS, Read NW 1984. Effect of painless rectal distension on gastrointestinal transit of solid meal. Dig. Dis. Sci. 29: 902-6.