EFFECTS OF HISTAMINE
AND PENTAGASTRIN ON FASTING EQUINE
GASTRIC AND DUODENAL CONTENTS
DIANE LYNN KITCHEN
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1997
Diane Lynn Kitchen
This dissertation is dedicated to my father, Hyram
Kitchen, D.V.M., Ph.D., who taught me the importance of the pursuit of excellence. I continuously strive to approach the high standards he spent a lifetime instilling in me. Even today, I look to his approval as the ultimate in reward. Thank you, Daddy, and I do wish you were here to share this with me.
I acknowledge the help of many individuals and animals who have been instrumental in my pursuit of this degree and the completion of this dissertation. I wish to thank Dr. Alfred M. Merritt, my mentor and the head of the Island Whirl Equine Colic Research Laboratory, where these studies were conducted. I thank him for the collaboration throughout the multiple attempts to define this project and various sidelines along the way, and also for the constructive criticism in the preparation of this dissertation and associated papers'.
I thank James A. Burrow for all the technical assistance over the years. His computer skills and laboratory assistance have been invaluable. Through the many changes in this project, the committee has remained intact and I thank them for staying on the committee. I thank Christie Stieble for her work on the statistical analysis of data for the balloon study. I thank Dr. Philip
Kosch for his role as Associate Dean for Research and Graduate Studies.
Dr. Martha Campbell-Thompson, who is one of the
pioneers in equine gastric secretion, provided a wealth of knowledge and was invaluable in the development of this project as did her chairman at the Health Sciences Center, Dr. J. McGuigan. I must also thank Dr. Dan Hogan and Mr. M. Koss who taught me not only the method for analyzing bicarbonate concentration, but who were instrumental in the construction of my specialized duodenal catheter.
For emotional support, I must thank several important people in my life. I am afraid that most have had to suffer me during one stage or another of this process and deserve accolades for this. I thank my mother, Yvonne H. Kitchen, R.N., for the shoulder. I know it has been a tough several years, but she were always there with unconditional love and I do love her for it. Michael S. Kitchen, M.D., my dearest brother keeps me humble. Dr. Jerry and Gayle Spears have had to "live" with me day by day and have always been there for me. I cannot thank them enough. Roger Reynolds gave me peace of mind and allowed me to focus on my own goals. Elmer and Harriet Heubeck, Cynthia McFarland, Gail
Leichliter and John R. Phillips have been loyal friends, providing me unending support throughout the course of this program. I also thank my private practice clients, who have had to deal with my erratic schedule and a goal they cannot comprehend.
A special acknowledgement and dedication must go to
Zero and One, two brave ponies who taught me so much in our brief acquaintance. The horses involved in these studies have each contributed not only data, but added to my life. I thank Adam, Buddy, Dick, Ethel, Harry, Iso, Jeff, Lucy, Mama, Spot, and Tom. Tejas, Clay, Blondie, and all my other four-legged family have provided my most ardent source of solace. A special thought for Valiant who did not make it to the end, but is still remembered fondly.
TABLE OF CONTENTS
ACKNOWLEDGMENTS . . . . . . . . . . iv
LIST OF TABLES . . . . . . . . . . xi
LIST OF FIGURES . . . . . . . . . . xii
ABSTRACT . . . . . . . . . . . xiii
1 INTRODUCTION AND REVIEW OF LITERATURE . . . 1
Introduction . . . . . . . . . . 1
Historical Background . . . . . . . 2
Review of Literature . . . . . . . . 5
Development of Hypothesis . . . . . . 1. 26
2 EQUINE HISTAMINE DOSE-RESPONSIVE GASTRIC ACID
SECRETION . . . . . . . . . 31
Introduction . . . . . . . . . . 31
Materials and Methods . . . . . . . 33
Results . . . . . . . . . . . 36
Basal Secretion . . . . . . . . 37
Pyrilamine Infusion . . . . . . . 39
Histamine dose-response . . . . . . 39
Volume . . . . . . . . . 39
Acid Concentration . . . . . . 40
Acid Output . . . . . . . 40
Sodium concentration . . . . . 41
Sodium Output . . . . . . . 42
Pentagastrin Outputs . . . . . . 42
Discussion . . . . . . . . . . 42
3 THE EFFECT OF PYRILAMINE MALEATE PRETREATMENT ON
PENTAGASTRIN- STIMULATED EQUINE GASTRIC
Materials and Methods.................54
Experimental preparation ............54
Experimental protocol. .............55
Analysis of data.................57
Results . . . . . . . . . . . 5
Post-Pyrilamine Collections ...........58
Early Infusion Collections...........58
Late Infusion Collections ............59
4 PLACEMENT OF A CATHETER FOR COLLECTION OF DUODENAL
CONTENTS IN HORSES WITH CHRONIC GASTRIC CANNIJLAS
AND ITS EFFECT ON GASTRIC SECRETION .. .....66
Materials and Methods.................68
Catheter Design............. ....72
Experimental protocol. .............74
Analysis of data.................77
Acid Concentration .............78
Sodium Concentration ............80
Sodium Output ...............80
5 THE EFFECT OF PYLORIC OBSTRUCTION ON EQUINE BASAL AND
STIMULATED GASTRIC SECRETION............86
Materials and Methods.................87
Analysis of data.................94
6 SUMMARY AND CONCLUSIONS...............130
A HISTAMINE DOSE-RESPONSE DATA.............138
B PYRILAMINE PRETREATMENT STUDY DATA..........144
C DATA FROM BALLOON/NO BALLOON STUDY..........147
D STATISTICAL ANALYSIS OF HISTAMINE DOSE RESPONSE
E STATISTICAL ANALYSIS OF PYRILAMINE MAIJEATE DATA .181
F STATISTICAL ANALYSIS OF CATHETER/NO CATHETER DATA 183
G STATISTICAL ANALYSIS OF BALLOON/NO BALLOON DATA .186
LIST OF REFERENCES . . . . . . . . . 202
BIOGRAPHICAL SKETCH . . . . . . . . . 213
LIST OF TABLES
1 Acid and Sodium Concentration Histamine Infusion 29 2 Equine Gastric Contents . . . . . . . 38
3 Acid and Sodium: Histamine versus Pentagastrin . 48 4 Gastric Contents with and without Pretreatment . 59 5 Horses used in Gastric Collection Studies . . . 68 6 Descriptions and Abbreviations for Each Study . 69 7 Data from Study With and Without Duodenal Catheter 79 8 Electrolyte Composition of Duodenal Contents . . 81 9 Experimental Design . . . . . . . . 91
10 Volume and Ph of Gastric Contents . . . . 99 11 Potassium in Gastric Contents . . . . . 101 12 Chloride in Gastric Contents . . . . . 103
13 Acid in the Gastric Contents . . . . . 106
14 Sodium in Gastric Contents . . . . . . 108
15 Bicarbonate in Duodenal Contents . . . . ill 16 Sodium in Duodenal Contents . . . . . 112
17 Potassium in Duodenal Contents . . . . . 113 18 Chloride in Duodenal Contents . . . . . 113 19 Volume of Duodenal Contents . . . . . 115
LIST OF FIGURES
1 Mean acid output/15 min. of each treatment. ... 45
2 Regression analysis of acid vs. sodium outputs during
pentagastrin and histamine infusion........49
3 Mean pentagastrin stimulated acid output with and
without pyrilamine pretreatment. ...........64
4 Cross sectional view of duodenal catheter passing
through the gastric cannula and to the duodenum . 71 5 Specialized catheter.................72
6 Specialized Balloon catheter............89
7 Cross sectional view showing balloon catheter
positioned in the proximal duodenum..........90 8 Acid concentration. A curvilinear line .......121 9 Acid output. A curvilinear line. ..........122
10 Sodium concentration. A curvilinear line . ..124 11 Sodium output. A curvilinear line. .........125
12 Collections from catheter or cannula. ........134
Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
EFFECTS OF HISTAMINE AND PENTAGASTRIN ON FASTING EQUINE
GASTRIC AND DUODENAL CONTENTS By
Diane Lynn Kitchen
Chairperson: Alfred M. Merritt, II Major Department: Veterinary Medicine
The composition of equine gastric contents has been
determined to differ markedly from that of other monogastric species. In the fasted animal, there is a voluminous sodium-rich fluid component which becomes greater during pentagastrin infusion, which is not eliminated by acid blockade. Potential pancreatic stimulation by pentagastrin was suggested by Alexander and Hickson. Histamine infusion, however, seemed to cause a classic parietal secretion, as opposed to the mixed parietal and nonparietal response to pentagastrin. These findings led to the formulation of the hypothesis for this dissertation: In the equine gastric
cannula model, the vigorous sodium-rich component of pentagastrin stimulated gastric contents is extragastric in origin.
The testing of this hypothesis was accomplished by a series of studies on horses with chronic gastric cannulae. A full histamine dose-response study was necessary to determine an optimal dose for the stimulation of maximal acid output. Since this required pretreatment of the horses with the H-1 receptor antagonist, pyrilamine maleate, a small study of the effect of this pretreatment on pentagastrin-stimulated secretion was designed. A technique was developed for the placement of an intraduodenal ballooned catheter via the gastric cannula to occlude the pylorus. A study was also performed to evaluate the effect of the presence of the intraduodenal catheter. Finally, the critical study involving a comparison of the gastric and duodenal contents during histamine and pentagastrin infusion, with and without obstruction of the pylorus, was done.
Gastric contents were analyzed in each of the studies for volume and electrolyte composition. The histamineinduced maximal acid ouput was equivalent to that previously
reported for horses stimulated with pentagastrin. Pyrilamine maleate dampened the gastric acid secretory response to pentagastrin infusion. The introduction of an intraduodenal catheter did not alter the maximal acid output to pentagstrin, although it may have enhanced the reflux of duodenal fluid into the gastric lumen. Most importantly, the obstruction of the pylorus with the balloon catheter significantly decreased the sodium output in gastric collections suggesting that the vigorous sodium-rich component of pentagastrin stimulated secretion is primarily of extragastric origin.
INTRODUCTION AND REVIEW OF LITERATURE
The regulation of gastric secretion has long interested scientists and clinicians due to the vital role gastric acid secretion plays in the normal gastrointestinal function and in certain pathologic conditions. The stomach, with both endocrine and exocrine function, is not a simple collection vat for food prior to digestion. A diverse group of cells with numerous capabilities are found within the highly
specialized gastric mucosa.1 Parietal cells secrete acid at the apical membranes containing the H+/K+ ATPase pump.2 Other cells are responsible for the production of mucus, zymogens, and other biologically active peptides and
amines.1 Numerous endocrine cells secrete substances that modulate various gastrointestinal functions.' The secretion of hydrochloric acid by the parietal cells is regulated by neural, hormonal, and paracrine mechanisms which interact to
control the rate of secretion.'1-3 Acetylcholine is the primary neurotransmitter involved in gastric acid secretion. Muscarinic nerve endings near the gastric glands release acetylcholine as part of the vagal reflex system.3 Enteric neurons in Meissner's plexus release GRP (gastrin releasing peptide) which acts on G cells in the gastric antrum. Gastrin is the peptide hormone released in response to specific substrates within the gastric lumen and distension of the antrum. Released by antral G cells, gastrin is an important hormone which is involved in all phases of gastric secretion and also exerts influence on DNA transcription and cellular replication.' Histamine acts directly on parietal cells, as a paracrine substance, to stimulate gastric acid secretion via H-2 receptors. Histamine originates from enterochromaff in-like (ECL) cells present within the mucosa, in response primarily to gastrin stimulation, with modulation by numerous other substances.'1
Understanding the mechanisms of gastric acid secretion and the regulation of these mechanisms has required decades
of research by many eminent scientists using a wide array of models and techniques. From the time of Pavlov and the fistulated dogs until the "in vitro" cellular studies of H~kanson et'al., debate has raged around the regulation of gastric acid secretion. Pavlov's work with fistulated models confirmed that subjective observations of a link between emotional status and digestive function were correct and introduced the role of a central neural component of
gastric secretory regulation.' Even today we use the term "Pavlovian" when referring to the brain's input into a function, such as the cephalic phase of gastric secretion.
Bayliss and Starling paved the way to investigation of hormonal substances involved in gastrointestinal secretion.5 In 1906, Edkins suggested that a hormone "gastrin"1 was released in the pyloric portion of the stomach and resulted
in the gastric phase of the acid secretory response.6 The presence of this distinct secretagogue was demonstrated by Komarov in 1938,7 and clarified in 1941 by Gregory and Ivy. B Gregory and Tracy were able to isolate the substance in two forms from pig antral mucosa and validate the hormonal regulation of gastric secretion by gastrin in the 1960s.9
Histamine was reported to stimulate the secretion of gastric acid in the early 1920s. Then, in the late 1930s, MacIntosh stated that vagal stimulation resulted in
histamine release within the gastric mucosa'0; however, controversy surrounded the role of histamine since known antihistamine agents were unable to prevent acid secretion."' The identification of different subtypes of histamine receptors by Black et al. in the 1970s and subsequent studies with H-2 receptor antagonist clearly demonstrated that histamine was an integral component in the normal gastric secretory response."," The complex interworking of each of these components confused investigators and resulted in conflicting studies. only in the 1970s, with studies focusing on "in vitro" cellular preparations and technical advances in immunology and histology, were scientists able to begin unravelling this complex regulatory process. The enterochromaff in-like (ECL) cell was the key to the combined action of acetylcholine,
gastrin, histamine and many other effectors.' Though ECL cells had been noticed in the early drawings and writings of Heidenhain, their role in the secretory response was not known.' The demonstration by HAkanson et al.'13 and Capella
et al.14 that the fundic ECL cells contained histamine, began the explosion of new concepts in gastric secretory regulation which has led to better understanding of the process.
Review of Literature
Today, it is apparent that gastric acid secretion by
parietal cells results from paracrine histamine released by nearby ECL cells. The release of histamine is mediated by many substances including gastrin, acetylcholine, somatostatin, vasoactive intestinal peptide (VIP), prostaglandins, calcitonin gene-related peptide, TGF-a, and even histamine.' Histamine acts on the basolateral membrane of parietal cells at H-2 receptors.23 Stimulation of these receptors activates the basolateral adenylate cyclase and leads to increased intracellular cAMP within the parietal cell."5 In contrast, parietal cell gastrin receptors and muscarinic M3 receptors result in increased intracellular Ca++ when stimulated.6 The increased cAMP and/or Ca++ may activate protein kinases involved in the phosphorylation of the apical H,K ATPase as occurs in Na,K ATPases.6 Several
cellular changes occur with activation of the secondary messenger system within the parietal cell. The proton pump, H,K ATPase, is present in the membrane of cytoplasmic membrane compartments during the resting or unstimulated state and, after stimulation within the membranes of extensive apical canaliculi. 16 The mechanics of this transformation are not clearly understood. It is possible that the cytoplasmic vesicles observed in the resting state are connected to the apical membrane and that tubules simply expand to expose the proton pump to the lumen of the gastric glands. An alternative explanation is that activation of the cell results in the fusion of cytoplasmic membranes with the apical membrane."'6
Another change occurring following stimulation of the parietal cell is the activation and transformation of parallel K and Cl channels, which provide the extracellular K+ needed for the proton pump.16 The pump requires oxygen as well as Mg+ The energy required by the proton pumps is supplied as ATP by the large mitochondrial content in the parietal cell.16 The proton pump transports H+into the gastric gland in exchange for K+. In order to maintain intracellular homeostasis, several membrane exchange
pathways are present within the membrane of the parietal
cell including Cl/HCO3 and Na/H exchangers.15 During stimulation, the parietal cell secretes H+ actively via the proton pump, Cl- passively via the Cl channel, and H,
passively into the gastric lumen.'5 Parietal secretion of HCl provides a high concentration (-140 mEq/L) of acid within the gastric glands that empty into the lumen of the stomach.
Gastric secretion is commonly described as either basal or stimulated. Basal secretion is also termed interdigestive secretion and occurs without external stimulation, such as ingestion of food, or conditioned response 2,1 In terms of volume and acid content, basal secretion varies greatly between species. '7 A circadian rhythm has been observed in human basal secretion with higher rates during the evening and low rates in the morning 2,3,17, 18 The rate of basal secretion is independent of serum gastrin concentration and is not abolished by vagotomy. 2.17
Stimulated gastric secretion is classically divided into three phases, with some overlap between them. These phases are commonly termed cephalic, gastric, and
intestinal.2,3, 118 The link between the brain and gastric function has been speculated about since the early 1800's.19 Sight, smell, taste, thought, and swallowing stimulate gastric acid secretion during the cephalic phase.2,39" The vagus nerve is the key link in the cephalic phase through its stimulation of gastrin release as well as direct innervation of the parietal cell.2,17"1 As mentioned, these effects may be mediated through ECL cells rather than directly through receptors on the parietal cell. Many neurotransmitters and certain peptides have been shown to act centrally on the regulation of acid secretion.9 Cholinergic agonists, prostaglandin inhibitors, GABAergic agonists, gastrin/CCK, TRH-related peptides, and somatostatin-related peptides administered into the CSF stimulate gastric acid secretion. Inhibition occurs following CSF injection of cholinergic antagonists, adrenergic agonists, prostaglandins, serotonin, bombesinlike peptides, opioid peptides, CRF-like peptides, and calcitonin-related peptides.19
The gastric phase of stimulated secretory response is a result of gastric distension or chemical stimulation of the gastric mucosa.2,3,17,18 This phase is mediated primarily
by gastrin. Distension of the stomach walls activates stretch receptors involved in release of gastrin. However, this release does not occur if the gastric contents are acidic due to a pH-sensitive, oxyntopyloric reflex with inhibition by somatostatin .217,118 Certain breakdown products of food are strong stimulants of gastric acid secretion, particularly peptides and amino acids, specifically, phenylalanine and tryptophan 2, "18 These stimulants induce the release of gastrin by action at the apical microvilli of G cells within the antral mucosa. 17
The intestinal phase of acid secretion is
responsible for only a small portion of the total acid secretory response *18 Stimulating -agents are the same as those involved in the gastric phase. The substances act directly at G cells in the upper small intestine as a gastrin-dependent mechanism and absorbed amino acids exert their effect by a gastrin-independent mechanism.' All phases of acid secretion are, in part, mediated by gastrin, through the release of histamine from ECL cells.'
The secretion of gastric acid follows the action of
histamine on H-2 receptors on the parietal cell. Inhibition of histamine release allows the parietal cell to return to a
resting state. Acidification of the gastric contents inhibits gastrin release by release of somatostatin from antral D cells."a The presence of acid, fat, or hyperosmolar solutions within the small intestine inhibits the acid secretory response through one or more of the peptides such as secretin, peptide YY, neurotensin, VIP (vasoactive intestinal peptide), GIP (gastric inhibitory peptide), and enteroglucagon released by cells within the small intestinal mucosa.1 011 These inhibitory substances have receptors on the ECL cell and have been shown to inhibit the release of histamine from these cells.' Histamine also exerts negative feedback on ECL cells via H-1 and H-3 receptors.' The rapid catabolism of histamine and its speedy removal from the extracellular space surrounding the parietal cell eliminates the stimulation of these cells and terminates the secretory process.20
Many drugs have demonstrated antisecretory actions that have been crucial in clarifying the mechanisms of acid secretion and in treatment of pathologic conditions related the secretion of gastric acid.21 Histamine H-2 receptor antagonists are able to prevent gastrin, food, and vagalinduced acid secretion. This finding was pivotal in the
determination of the central role of histamine in the secretory response.22 As a result, numerous H-2 blockers have become a mainstay in the treatment of peptic ulcers.21 Substituted benzimidazoles, such as omeprazole, have more recently joined the battery of antiulcer medication due to their inhibition of the H,K ATPase.' Certain prostaglandin E2 analogues also have an antisecretory effect, due to actions at CNS as well as ECL cell level."121 Gastrin receptor antagonists (ie. Proglumide) are also capable of inhibiting acid secretion.2' The target of some presently available drugs and many of the drugs of the future may not be parietal cells directly, but other cells within the gastric mucosa which act on the parietal cell.23 Accordingly, the increasing understanding of the role of the ECL cell is likely to make it the target of investigational agents for the regulation of acid secretion.
The gastric ECL cell is considered to be one of the amine precursor uptake decarboxylation (APUD) series of cells."24 Numerous endocrine cells are found scattered within the gastric mucosa and may represent from 0.5% to 2% of the cells present. The prevalence of ECL cells vary among species, as they are particularly common in the rat
and appear in lesser numbers in the dog and primates.1,24 Other endocrine cells observed in the gastric mucosa include enterochromaffin cells which produce 5-hydroxytryptamine (serotonin), D cells that secrete somatostatin, gastrinproducing G cells, and P and X cells which have unknown functions.' The ECL cells are found near parietal cells within the gastric glands and have prominent cytoplasmic extensions."24 They contain a large number of cytoplasmic vesicles and electron-dense granules, and take up aromatic amino acids and decarboxylate them. Histidine decarboxylase (HDC) and histamine are also found within these cells."124 Histamine is concentrated into cytoplasmic vesicles by a Vtype ATPase which acts as a H+/histamine antiporter.122,24 As indicated above, the production and storage of histamine by the ECL cell makes it the critical interface in the regulation of gastric acid secretion.124 The primary mediator of histamine release is gastrin acting through the CCK-B receptor to increase cytosolic Ca++.1.22,24 Histamine release can also be stimulated by acetylcholine at muscarinic M1 receptors, epinephrine through P adrenergic receptors, isoproterenol, forskolin, interleukin 18 and VIP,
and can be inhibited by somatostatin (subtype 2), TGF-, TGF-P, calcitonin gene-related peptide, and histamine via H1 and H-3 receptors.1222425 It appears that there is substantial neurotransmitter modulation of gastrin and its regulation of histamine release from ECL cells.25 Although, it is known that Ca++ acts as a secondary messenger within the ECL cell, cAMP is likely to also play a role in the release of histamine.12224
In addition to stimulating the release of histamine, gastrin regulates HDC activity and ECL cell proliferation."124 It stimulates DNA synthesis by a pathway distinct from that of histamine release, although both processes appears to begin with the binding of gastrin at the CCK-B receptors."124 Hypergastremia has been associated with increased numbers of ECL cells and may play a significant part in the development of gastric carcinoids."124 The chronic use of antiulcer medication with resultant hypergastremia have been shown to lead to ECL hyperplasia and gastric mucosal hypertrophy.122,24 The recent advances in the "in vitro" cellular techniques have expanded our knowledge of the regulation of the gastric secretory response; however, as history shows us, advances
just open the door to a multitude of additional questions. Although, much knowledge has been gained by isolated cell work such as that with the ECL cell, the wide array of species differences points to the continuing need for both "in vivo,, and "in vitro', studies.
Early classical studies were "in vivo," performed
primarily on animals with a variety of surgically prepared pouches or fistulas. Even human studies in the early 1900's, were performed on subjects with gastric fistulas.19 Historically, experimental preps commonly utilized dogs and cats with transplanted pouches or surgically isolated stomachs 8,26,27 over time, chronic gastric cannulas have been placed in the stomachs of dogs, rats, rabbits, pigs, cats, humans, and horses to allow studies to be performed on conscious animals 5.28-31 Amphibian mucosa was the first tissue used in the "in vitro" studies that provided the cardinal information about the acid transporting ATPase .32 Subsequent cellular studies have been performed on canine,
equine, rabbit, and rat gastric glands. Initially,
food and food extracts were selected to stimulate gastric secretion, later histamine and alcohol were used in many studies 8,26,27 Histamine continues to be a valuable
secretagogue for investigation of the acid secretory process and its regulation. 120,29,30 After Gregory and Tracy found that an extract of porcine antral mucosa was an effective stimulant of acid secretion9, gastrin became a popular investigational tool.28 A synthetic polypeptide analog to the C terminal tetrapeptide of gastrin is pentagastrin, or peptavlon, which has become the secretagogue most commonly used for stimulation of gastrin related secretion."'-,"1,30,31 Today, gastric secretory studies range from chronically implanted gastric cannulas in intact animal models to receptor immunochemical studies on ECL cells isolated from the gastric mucosa.
The hundreds of studies of gastric acid secretion have accumulated a vast amount of information with marked similarities in the process and its regulation. However, some striking differences have also been observed. The relative sensitivity to specific secretagogue, the rate of basal secretion, the response to inhibition, and ECL cell shape and prevalence are markedly different between species." In the rat, the gastric acid secretory process is highly sensitive to pentagastrin and very resistant to histamine .17,30 Histamine and gastrin are equally effective
on canine mucosa; 17,30 in rabbits histamine is much more effective than gastrin. 17,30 Basal secretion is minimal in dogs and cats and high in rats and rabbits, whereas humans, primates, pigs, and horses fall somewhere in between. 17,28-31 Maximal acid outputs on a normalized bodyweight basis induced by either histamine or pentagastrin, also differ between species. Maximal histamine-stimulated output is greater in rabbits than dogs, while in dogs it is greater than that of rats, pigs, and cats, which in turn, have rates of secretion markedly greater than that of primates and man. 1729'30 During pentagastrin-induced acid secretion, rabbit and dog have a comparable maximal response, but that of rats is significantly less 30 As with histamine, pig,' rat, and cat maximal acid secretory rates in response to pentagastrin are comparable .29337,38 Maximal acid output is used to determine the necessary dosage of either secretagogue for the purpose of most experimental work; however, the gastric contents collected during these studies consist of fluid containing many electrolytes, pepsin, mucus and other substances, as well as acid.
The electrolyte composition of gastric secretion other than H+ also changes with the secretory rate. The secretory
pattern has been examined with both histamine and pentagastrin in experimental models394 and in patients with gastroduodenal diseases.42 In man, the infusion of either histamine or pentagastrin results in similar changes in gastric electrolyte concentrations and outputs. Sodium concentration decreases and output remains constant, while potassium concentration increases slightly and output increases significantly. The concentration of calcium and phosphate decreases and outputs are unchanged.39 Acid and sodium concentrations have a strong inverse correlation under these stimulatory conditions.39 The histaminestimulated maximal acid output in dogs is greater than that of pentagastrin; however, peak acid output is reached more quickly with pentagastrin than with histamine.41 Sodium ion concentration is inverse to acid in the canine studies just as in human studies.
Thus, gastric secretions are a combination of parietal and nonparietal fluids. The parietal component is isotonic HCl that is secreted in response to specific secretagogues at a rate that increases significantly with simulation of the parietal cell.' The nonparietal component is believed to be a constantly secreted isotonic ultrafiltrate of
plasma."3 No change in nonparietal secretory rate is observed during stimulation with any of the known gastric secretagogues. Sodium ions are the predominate cation of this nonparietal component. The differences in response to secretagogues of these two components explains the inverse relationship of sodium and acid concentrations reported during increasing acid secretory rates.43
Various gastroduodenal diseases are characterized by alterations in the electrolyte composition of gastric contents.4 Patients with active duodenal ulcers have increased acid and chloride concentration and decreased sodium concentration, while gastric ulcer patients have increased sodium concentrations in-the gastric content without changes in the concentrations of acid, potassium, or chloride.4 Other changes in the composition of electrolytes are suggestive of trophic gastric ulcers, chronic gastritis, and Zollinger-Ellison syndrome."' The characteristic gastric secretory response to stimulated gastric secretion has been consistently observed in all species except the horse .4
Due to the relative difficulty in collecting the
contents, little was known about equine gastric secretion
until the 1980s. The apparent increase in gastric and duodenal ulcer disease in horses provoked a particular interest in'equ~ine gastric physiology.11'44 Gastric mucosal ulceration has been observed in horses as a sequela to the administration of nonsteroidal anti-inflammatory drugs, and as a spontaneously occurring problem of unknown etiology. 44-47 Intermittent aspiration of gastric contents via nasogastric tube, placement of indwelling pH probes nasogastrically, and chronic indwelling gastric cannulas to collect contents by drainage have been used in recent studies of equine gastric secretion.'485 Only the cannula allows complete collection of gastric contents, but these studies can only be performed on fasted animals. Intermittent aspiration also requires fasted horses, but may not collect gastric contents in their entirety. The indwelling pH probe makes it possible to monitor the intragastric environment for a long periods of time in either fed or fasted states.
Basal secretion in horses has been described as continuously variable, with periods of spontaneous alkalinization as has also been observed in humans, pigs, rodents, monkeys, and chickens.14'85 Thus, the hydrogen ion concentration and acid output range widely during
unstimulated gastric collection periods .31,4 The basal equine acid output of approximately 30% of maximal output is consistent with pigs and rats, and it is greater than that of humans.44 On a body weight basis, equine basal acid output over time is similar to that found in man.34
Up to now, pentagastrin has been the only
secretagogue used "in viva" to stimulate equine gastric acid secretion to characterize the species-specific responses and to study drugs designed to prevent or treat gastric ulcers. 48,49,51,53,54 Its administration to horses results in a significant increases in both gastric acid and sodium outputs.44 Acid concentration increases after stimulation, but not to the maximal magnitude observed in other species. In addition, the increased sodium output during stimulation is unique to horses. As the dose of pentagastrin is increased, the concentration of sodium decreases slowly and remains greater than the acid concentration, unlike in other species where the sodium concentration rapidly decreases in an inversely proportional relationship to the increasing acid concentration .44 This is a major species-specific difference in the equine secretory response to pentagastrin. Thus, it has been suggested pentagastrin stimulates both the
parietal and nonparietal components of gastric secretion in the horse.44 A variety of compounds have been shown to inhibit equine basal and pentagastrin-stimulated gastric acid secretion, such as histamine H-2 antagonist ranitidine, the proton pump blocking agent omeprazole, and the prostaglandin misoprostol,48495153,54 The profound sodium rich nonparietal secretion characteristic of the equine response to pentagastrin is evident even after omeprazole has effectively inhibited the acid secretory response.5' The origin of this voluminous nonparietal component of gastric contents has not been clearly defined.
The periods of high pH of gastric contents during basal secretion have raised much speculation about the equine gastric secretory capacity, gastric emptying rates, and the buffering capacity of the various secretory products.50 Potential buffering fluids include saliva, gastric nonparietal secretions, and duodenogastric reflux.44 Saliva is not a likely candidate since the production of saliva in the horse is minimal except during mastication.55 Fasting equine gastric contents are frequently dark green to yellow in color and viscous, suggesting possible contamination with bile.44 Duodenogastric reflux has been observed during
endoscopic examination of the stomach."1 The fluid refluxed could potentially contain a mixture of biliary, pancreatic and duodenal secretions, since the duodenal diverticulum, where the bile and minor pancreatic ducts enter, is in relatively close proximity to the pylorus.51 Furthermore, the absence of a gallbladder leads to continuous secretion of bile in horses and equine pancreatic secretion is reported to be profuse and continuous. ,56
Pancreatic secretion of the horse is distinct from that of other species. The resting secretion of approximately 25 l/g gland/min is increased by four to fivefold during eating, stimulation of the vagus nerve, or after the administration of secretin or pentagastrin.55,7 Neither resting nor vagal-induced pancreatic secretion is abolished by atropine.55,58 Vagotomy does not alter the basal secretion. Studies with ponies fitted with re-entrant cannula in the pancreatic duct found that they may secrete a volume equivalent to up to 10% of their own body weight daily.55 The concentration of amylase and the proteolytic activity of equine pancreatic juice is particularly low, and it has a very weak ability to emulsify fat, compared to other species.." Small increases in the concentration of
amylase are found during stimulation of the vagus, or during secretin or pentagastrin administration. The concentration of sodium and potassium are similar to respective plasma concentrations and chloride is the predominate anion at all levels of secretion. The bicarbonate concentration is low relative to other species and does not increase to the extent anticipated when the volume of secretion increases.55 That is, in most species, the concentration of bicarbonate is less than chloride concentration at rest, but increases to become the predominate anion during maximally stimulated secretion .55 The composition of the large non-parietal component of the fasting gastric contents is also similar to that of equine pancreatic fluid. Therefore, the observed species-specific particulars of equine pancreatic juice contents might explain the voluminous fluid response observed during pentagastrin-stimulated gastric collection.
The pancreas is a multifunctional organ which is
responsible for the secretion of fluid, electrolytes, and enzymes essential to the normal digestive process and the endocrine secretion of hormones critical to metabolic homeostasis .59 Just as with gastric secretion, pancreatic secretion has traditionally been divided into basal and
meal-induced patterns, with meal related secretion further divided into cephalic, gastric, and intestinal phases.60 In most species, basal enzyme secretion ranges from 10% to 30% of maximal secretion, whereas, basal bicarbonate secretion is only 1% to 2% of maximal.60 The regulation of basal enzyme secretion appears to vary between species, while basal bicarbonate secretion is dependent primarily on secretin augmented by cholinergic neural input.60 Basal enzyme secretion is almost entirely due to cholinergic stimulation in the rat, whereas, in man, both CCK and cholinergic stimulation mediate it. Brief bursts of pancreatic enzyme and bicarbonate secretion are associated with the migrating myoelectric complex and a circadian rhythm has also been reported.60 The normal pancreatic secretory response to a meal is increased enzymatic secretion to aid digestion and increased bicarbonate secretion to buffer the chyme and assure the optimum intraluminal pH for enzymatic activity. As with gastric secretion, "cephalic" pancreatic secretion seems to be stimulated either directly or indirectly by the vagus nerve.50 "Gastro-pancreatic reflexes" are initiated by food or gastric distension and mediated by cholinergic vagovagal
reflex.50 The two classical pancreatic regulatory hormones, cholecystokinin (CCK) and secretin, and vagal cholinergic reflexes are involved in the "intestinal" phase.0 Postganglionic cholinergic neurons are important in the regulation of enzyme and bicarbonate secretion by the release of acetylcholine at muscarinic receptors. Secretin is the most potent stimulant of the bicarbonate-rich fluid component of exocrine pancreatic secretion in dogs, cats, rats, and humans., while the primary hormonal stimulant of enzyme secretion is CCK.9,60 In the rat, rabbit, pig and guinea pig, CCK is responsible for a copious fluid and electrolyte secretion which may have a low concentration of bicarbonate and a high concentration of chloride.59 Enzyme secretion is stimulated by the action of CCK at CCK-A receptors on pancreatic acinar cells.61 Although CCKB/gastrin receptors have been identified on pancreatic acinar cells, their function remains unknown.61 Caerulein is even more potent than CCK, and both are significantly more potent than gastrin, at stimulating pancreatic enzyme and fluid secretion in dogs.62 A multitude of other receptors are present on pancreatic acinar cells, including those to bombesins, tachykinins, VIP, somatostatin, insulin,
endothelin, insulin-like growth factor and epidermal growth factor. Much of the electrolyte secretion by the pancreas is secreted by duct cells rather than acinar cells 63 The electrolyte composition of pancreatic secretions vary with the rate. Bicarbonate and chloride concentrations have a reciprocal relationship, but the cumulative anion
concentration remains constant. As the secretory rate increases, bicarbonate concentrations increase and chloride concentrations decrease .63 The volume of the secretion in response to secretin vary with species. Cats, dogs, pigs and man have greater secretory rates per gram of gland than
rats and rabbits. The specific differences observed in the limited studies on equine pancreatic secretion are unique to the horse and similar secretory patterns have not been reported in other species.
Development of Hypothesis
After reviewing the overwhelming information about
gastric secretory physiology in general, and equine gastric secretion specifically, it is apparent that equine gastric physiology requires still more investigation. The
development of a silastic indwelling gastric cannula by Campbell-Thompson and Merritt for the collection of equine gastric secretions was pivotal. The distinct composition of equine gastric contents, particularly the uncharacteristic "non-parietal" response to pentagastrin stimulation, is of special interest. Since pentagastrin has been the only exogenous secretagogue utilized to date in equine secretory studies, it would be of interest to see if the horse responded to other secretagogues in a similar way. Therefore, the initial project for this series of studies was to characterize the equine gastric secretory response to histamine. Pilot studies were performed with moderate trepidation due to the potential side effects of histamine administration to horses. These studies required pretreatment of the horses with a H-1 antagonist, pyrilamine maleate. The results of a small pilot study were the last key to the development of a hypothesis for this dissertation.
Histamine-stimulated equine gastric contents were markedly different than those following pentagastrin stimulation reported in previous studies. The mean acid and sodium concentrations from the pilot study are presented
in Table 1. It was apparent from these studies that the horse is capable of developing the "classical" parietal response to an acid secretagogue that is comparable to that of other monogastric species. During histamine stimulation, the maximal acid concentration values were much greater than those induced by pentagastrin, and sodium concentrations decreased in the classical reciprocal relationship to acid. The sodium output over various doses of histamine was constant, unlike the pentagastrin responses. Yet, the maximal acid output provoked by histamine was equivalent to that of pentagastrin stimulated secretion. These pilot findings made it evident that the stimulation of gastric acid secretion with histamine could be safely performed in horses, with certain precautions, and that well-designed dose response study was needed. This led to the study described in chapter 2. The predominant conclusion of early investigation of histamine stimulation was that, as opposed to pentagastrin, it induced a purely parietal gastric secretion in horses.
Acid and Sodium Concentration during Histamine Infusion (n=4 horses)
Histamine Mean [H+] Mean [Na+]
Dose mEq/L mEq/L
Basal 26.5 101.3
15 73.9 57.8
30 95.1 55.5
45 97.5 52.4
In order to perform controlled comparisons between histamine and pentagastrin, without potential extraneous factors, it was necessary to evaluate the secretory response to pentagastrin following pyrilamine maleate pretreatment. Chapter 3 presents the unexpected results of that study.
Four crucial facets of equine gastric secretion were considered to develop the dissertation hypothesis. First, equine basal and pentagastrin-stimulated gastric secretion has an uncharacteristically low concentration of acid and high concentration and output of sodium. Second, acid blockade with numerous agents does not eliminate the profound sodium-rich nonparietal component of pentagastrinstimulated secretion. Third, Hickson and Alexander reported
that equine pancreatic water and electrolyte secretion is profuse and continuous and can be stimulated by pentagastrin. Finally, the stimulation of gastric secretion with histamine resulted in purely parietal secretion resembling that of other species. Therefore, the dissertation hypothesis is as follows. In the equine gastric cannula model, the vigorous sodium-rich component of pentagastrin stimulated gastric contents is extragastric in origin.
A technique of duodenal catheterization through the
chronic gastric cannula was developed to allow for occlusion of the pylorus during acid secretory studies. Chapter 4 discusses the technique for duodenal catheterization, the effect of the technique on gastric secretion, and the composition of duodenal contents. The primary study of this dissertation, describing results with and without the obstruction of the pylorus, is presented in chapter 5 and was designed to definitively determine if the sodium rich fluid found during pentagastrin stimulation and absent during histamine stimulation is secreted by the gastric glandular mucosa or by a extragastric source.
EQUINE HISTAMINE DOSE-RESPONSIVE GASTRIC ACID SECRETION
Previous studies of equine gastric physiology have all utilized pentagastrin to stimulate gastric secretion. 44, 50,52 Increased recognition of clinical disease in horses due to gastric and duodenal ulceration has led to interest in the potential use of horses as a model for peptic ulcer disease in humans as well as a desire to increase our knowledge about horses, themselves. Gastric- ulceration has been shown to be a widespread phenomenon in horses and foals .64 Compared to other species, pentagastrin-stimulated gastric contents in horses differs by being relatively low in acid concentration and high in sodium concentration, even at maximal secretory responses .44 Nevertheless, the inhibition of pentagastrin-stimulated gastric acid secretion has been an important and effective means to evaluate therapeutic potential of various anti-ulcer agents in this specie S. 31,49 31
Histamine-2 receptor antagonists have been used successfully to treat horses with clinical signs related to gastric ulceration. 64
In order to better characterize equine gastric
physiology, investigation of the effects of a secretagogue other than pentagastrin is essential. Histamine seems the obvious choice, but its use in horses has been avoided due to the presumed equine respiratory and CNS hypersensitivity to this agent. Ideally, "Histalog", a specific H-2 agonist, might be used, but it is no longer available. In other species, the undesirable side-effects of histamine have been largely avoided by pretreatment with an H-1 antagonist29, and such a protocol was chosen for the studies described herein. Specific objectives of the study were to: 1) determine the dose of histamine needed to elicit a maximal acid secretory response in the horse and how equidae compare with other species, and; 2) see if the non-parietal component of the secretory response to histamine is similar to, or different from, that seen with pentagastrin stimulation.
Materials and Methods
Six adult horses, both mares and geldings, were used.
All studies were done with the approval of the University of Florida IACUC. The horses ranged from 2 to 20 years of age. The five Thoroughbreds, and one Arabian weighed an average of 484 kg (range 444-506 kg). All horses were free of clinical signs of gastrointestinal disease, were deformed every 2 months and were vaccinated for encephalitis and tetanus every 6 months. They were maintained on grass pasture with coastal hay ad liband 12% protein grain twice daily. Each horse was previously prepared with a chronic indwelling gastric cannula as described by Campbell-Thompson and Merritt .31
The horses were fasted with free choice water for 20
hours prior to each experiment. At least one week interval occurred between experiments on any given horse. Studies were performed while the horse was loosely restrained in the laboratory.
After placement of an indwelling jugular catheter, the gastric cannula was opened and drained by gravity for 30
minutes allowing emptying of residual gastric contents. During the experiments, gastric contents were collected in 15 minute aliquots. The volume was measured to the nearest
5 ml in a graduated cylinder. Samples were filtered through gauze to remove feed particles and foam. The pH was determined using a glass electrode (Radiometer, Copenhagen,
Denmark) calibrated at 200C using commercial buffer solutions of pH 2 and 7 (pH standard, Fisher Scientific). Hydrogen ion concentration was measured in duplicate by electrometric titration with O.1N NaOH to an endpoint of
7.4. (Radiometer, Copenhagen, Denmark) Sodium ion concentration was measured by flame photometry (Instrumentation Laboratories Inc., Lexington, MA) on samples diluted in an internal standard lithium solution (Dilumat, Fisher Scientific). The machine was calibrated
with known [Na+]/[K-1] standards (Instrumentation Laboratories Inc., Lexington, MA) prior to analysis and after every 5 samples. Acid and sodium outputs were calculated for each time period on a per kg body weight basis.
The first three 15-minute time periods were during basal (no treatment) gastric collections. At time t=45
minutes, pyrilamine maleate ("Histavet-P", Schering-Plough, NJ) was infused IV at 1 mg/kg over the entire 15 minute collection period. No treatment was given during the subsequent two collection periods (t=60-90 minutes). Histamine infusion (7.5 pg/kg-hr) was begun at time t=90 minutes and continued for 60 minutes. Two additional 60 minute infusions of histamine, of 15 jg/kg-hr and 30 pg/kg-hr respectively, followed. Therefore, the whole experiment lasted for 4.5 hours.
Crystalline histamine (Sigma Chemical Co., St. Louis Mo) was dissolved in 60 mls of 0.9% NaCl and filtered through a 0.22 pm cellulose nitrate filter (Corning, Corning NY) in preparation for infusion, given by infusion pump. (Harvard Apparatus, South Natick MA)
In a separate trial, the horses were maximally
stimulated with pentagastrin44 for collection of data for comparison. The horses were prepared as described for histamine trial. After two hours of basal collection, pentagastrin was infused at 6 jg/kg-hr for two hours. Gastric collections were analyzed as during the histamine trials.
Maximal acid output per histamine dose was calculated
from the last two 15-minute collections at each dose. Values were expressed as pEq/kg BW/15 min. Results are presented as mean and SEM for all parameters. Data were subjected to one-way analysis of variance for repeated measures. A significance value of p<0.05 was selected. Pairwise multiple comparison testing of significance was performed using Student-Newman-Keuls test. Acid output vs. sodium output data were subjected to linear regression for comparison to respective pentagastrin-stimulated outputs from the same horses.
ResultsThe horses used had been involved in gastric secretory studies for at least one year prior to these studies. They demonstrated no evidence of gastrointestinal disease or problems related to the gastric cannula. Endoscopic examination of the gastric and duodenal mucosa as a part of another project in the laboratory revealed an occasional Gastrophilus spp. larvae and healthy intact gastric squamous and glandular mucosas.
Two horses did not complete the entire experiment. In one experiment, the infusion had to be terminated after the 15 .g/kg-hr histamine infusion period because the horse developed marked muscle tremors and mild abdominal contractions. After the infusion was stopped, the gastric cannula closed and hand-feeding begun, the horse recovered rapidly. Thirty minutes into the 30 tg/kg-hr infusion into a second horse, it developed generalized muscle tremors and mild synchronous diaphragmatic flutter. It became clinically normal within 20 minutes of ending the experiment and being allowed to eat.
Contents collected under pretreatment (basal)
conditions were generally yellow-tinged, viscid, and cloudy. The 15-minute aliquots ranged in volume from 280 to 640 mls, with a mean of 428.3 + 34.1 mls/15 min. The mean acid concentration was 42.5 + 3.9 mEq/L, and mean acid output was 39.8 + 6.6 JLEq/kg BW/15 min with a range from 19.3 to 82.7 pEq/kg/15 min. The pH of the samples varied greatly between 1.38 and 2.01, with a mean of 1.67 + 0.036. The mean sodium concentration was 75.7 + 7.2 mEq/L, and mean sodium
output was 68.2 + 8.2 pEq/kg BW/15 min with a range from
17.7 to 101.7 iEq/kg/15 min.
Data from Equine Gastric Contents Collected Before and After Histamine Infusion.
BASAL POST- 7.5 15 30
PYRIL- (pg/kg- (pg/kg- (pg/kgAMINE hr) hr) hr)
Volume 428.3 + 356.7 + 591.7 + 591.7 + 508.9 +
(mls/ 34.1 37.3 41.4 28 54.8
15 min) a,b a,b a,b
pH 1.67 + 1.79 + 1.46 + 1.48 + 1.29 +
0.036 0.08 0.09 0.13 0.03
Peak 42.5 + 33.5 + 72.2 + 81,3 + 82.9 +
[Hi] 3.9 3.5 3.9 5.1 6.8
(mEq/1) a,b a,b,c a,b,c
PeakAcid 39.8 + 25.9 + 87.9 + 99.5 + 98.4 + Output 6.6 4.4 7.7 8.1 4.2
(pEq/kg/ a,b a,b,c a,b,c
Peak 75.7 + 79.5 + 48.8 + 46.7 + 39.1 +
[Na'] 7.2 7.5 5.8 6.2 9.0
(mEq/1) a,b a,b a,b
Peak Na 68.2 + 60.6 + 61.7 + 58.5 + 50.3 + Output 8.2 9.6 8.8 9.4 14.7
Values expressed as Mean + SEM Pairwise Multiple Comparison procedure-Student-Newman-Keuls
a-significantly different (p<0.05) than basal
b-significantly different (p<0.05) than post-pyrilamine
c-significantly different (p<0.05) than 7.5 pg/kg-hr Peak values derived from the final 30 minutes of each treatment.
The volume of contents collected ranged from 70 to 550 mls, with a mean of 356.7 + 37.3 mls/15 min and was not significantly different than basal collections. The mean acid concentration was 33.5 + 3.5 mEq/L, mean sodium concentration was 79.5 + 7.5 mEq/L, and mean sodium output was 60.6 9.6 pEq/kg BW/15 min. Acid output ranged from
3 to 55 tEq/kg BW/15 min, with a mean of 25.9 + 4.4 pEq/kg BW/15 min. These results were not significantly different from the basal time periods.
Gastric collections became progressively more clear as acid concentration increased. During maximal histamine stimulation, the contents were colorless and watery. Volume
Maximal secretory volumes ranged from 360 to 850
ml/15min. For each increasing dose of histamine, the mean volume was 591.7, 591.7, and 508.9 ml/15min, respectively. The volume collected during each dose of histamine infusion was significantly greater (p<0.0001) than basal volume;
however, the stimulated volumes did not differ significantly among doses.
The mean acid concentration during basal and increasing histamine infusion doses showed the expected vigorous response to the treatments. Maximal acid concentration (MAC) ranged from 51 to 110 mEq/L, and mean MAC was significantly greater (p<0.0001) than basal acid concentration at all doses. Acid concentration at all doses was also significantly greater (p<0.05) than during the
post-pyrilamine period. The [H+1 during the 7.5 jg/kg-hr infusion was significantly lower (p<~0.05) than during either the 15 or 30 jg/kg-hr infusion. The MAC during 15 and 30 jg/kg-hr infusions did not differ significantly. Acid Output
Acid output (AC) for each treatment was calculated as the mean of the final two 15 minute collections of each infusion period. The individual highest AC was 155 jEq/kg BW/15 min occurring at t=195 min during the 15 jg/kg-hr histamine infusion. The AC in response to all histamine doses was significantly greater (p<.0001) than basal
output, and was the greatest during the 15 pg/kg-hr infusion. However, there was no significant difference between the responses to 15 pg/kg-hr and 30 pg/kg-hr infusions, whereas both the 15 and 30 pg/kg-hr infusion resulted in an AO significantly greater (p<0.05) than that from the 7.5 pg/kg-hr infusion. Sodium concentration
The sodium concentration was inversely related to acid concentration and ranged from 48.8 mEq/L to 39.1 mEq/L. The mean sodium concentration was significantly less (p=0.000119) during the 15 and 30 pg/kg-hr histamine infusion than during the basal period. The 7.5 pg/kg-hr infusion resulted in (Na] significantly greater than either the 15 or 30 pg/kg-hr infusion and not significantly different from either basal or post-pyrilamine periods. The concentration of Na in collections during the 15 and 30 pg/kg-hr infusions did not differ significantly from one another.
Mean maximal sodium outputs decreased from 61.7 +
8.8 jEq/kg/15 min to 50.3 + 14.7 pEq/kg/15 min as the histamine infusion rates increased from 7.5 Jtg/kg-hr to 30 pg/kg-hr. No significant differences were observed in the sodium output during basal, post-pyrilamine or histaminestimulated collection periods. Pentagastrin Outputs
The mean maximal acid output during pentagastrin
infusion was 91.8 + 3.5 gEq/kg/15 min. Corresponding mean peak sodium output during the pentagastrin trial was 100.8 +
5.0 pEq/kg/15 min.
As in other species, intravenous administration of
histamine base stimulated gastric acid secretion in horses resulting in maximal acid outputs (MAO) comparable to those of pentagastrin-stimulated horses. Infusions were performed without serious complications. Pretreatment with pyrilamine
maleate, a selective H-i receptor blocking agent", presumably eliminated any potential respiratory or neurologic side-effects of histamine infusion. Histamine-induced dyspnea results from bronchoconstriction mediated by H-I receptors." Both H-i and H-2 receptors are located in the central nervous system and horses are more susceptible to pronounced excitation and anxiety than some other species. The histamine provocation test for studies of equine bronchial responsiveness requires sedation to prevent apprehension and anxiety.65
Two experiments were not completed because the horses
developed synchronous diaphragmatic flutter (SDF) and muscle twitches. It is our belief that this was due to metabolic alkalosis with hypocalcemia and hypochloremia following marked HCl secretion.6 Changes in the ionization potential during metabolic alkalosis alter the free to bound calcium ratio and results in SDF.66 Once stimulation of HCI secretion ceased and the cannula was occluded, fluid and electrolyte loss ended and the horses responded rapidly. These individual horses were apparently hyper-responsive to histamine, since the MAO was reached during the 7.5 [tg/kg-hr infusion in one horse and after the first 15
minutes of the 15 ptg/kg-hr infusion in the other. These horses did not exhibit long-term effects following these episodes, and subsequent histamine infusions were administered to them without incident.
Antihistaminic agents, such as pyrilamine maleate, specific to H-I receptors are not expected to have antisecretory effects and are routinely used as part of the histamine challenge of gastric secretion in humans." A post-pyrilamine decrease in acid concentration and output was consistently observed and could reflect receptor cross-reaction or changes in blood flow to the gastric mucosa, since both H-1 and H-2 receptors are involved in histamine effects on vasculature.11 H-1 receptor blockade may also affect the delivery of substances to the basolateral membrane of parietal cells, since H-i receptors are involved in the regulation of capillary permeability."
The results indicate that a histamine dose of 30
jLg/kg-hr can be considered as that which will induce a maximal gastric acid secretory response in the horse. (See fig.l) In pilot studies for this trial, horses were infused with 15, 30, and 45 gg/kg-hr of histamine. However, some of these horses exhibited signs of metabolic alkalosis
BASAL PYRILAMINE 7.5 15 30
Figure 1. Mean acid output/15 min. during the last 30 minutes of each treatment. "Basal" indicates the status prior to any treatment. "Pyrilamine" indicates the output after treatment with pyrilamine maleate (1 mg/kg IV) and peak outputs during IV histamine infusion of 7.5, 15, and 30 pg/kg-hr are indicated accordingly. Data are expressed as mean + SEM.
during the 45 pg/kg-hr infusion; therefore, we elected not to utilize this higher dose in this trial. Four horses completed the pilot study and the MAO during the 45 g/kg-hr infusion was not significantly different from the MAO during
30 [tg/kg-hr infusion in these horses.(unpublished results) As with other species2941, some individual horses were maximally stimulated at lower doses, but there was no "supramaximal depression,40,67 of the mean response seen when the dose was increased from 15 to 30 gg/kg-hr. Thus, the horse is apparently slightly more sensitive to histamine than man, in which the infusion of 42 jg/kg-hr results in MAO(1); the pig and dog are more resistant, requiring 60 jg/kg-hr29 and 50 pg/kg-hr24, respectively. The peak secretory response to histamine in horses was more gradually attained than the response to pentagastrin infusion.44 This phenomenon has been previously reported in other species, as well.41
Mean maximal acid output in horses induced by histamine was similar to that seen in humans on mEq/kg basis. Furthermore, the maximal responses to histamine and pentagastrin are not significantly different in horses, as was anticipated. Species differences in maximal acid output and relative sensitivity to histamine and pentagastrin are numerous.17,8 The majority of "in vivo" gastric secretory studies have involved dogs, rats, and humans.69 Dogs have
minimal basal acid secretion, whereas humans and rats have active, though erratic, basal secretion, as do horses.44 The maximal acid secretory response to histamine and pentagastrin is equivalent in dogs and humans, while rats are much more sensitive to pentagastrin than histamine. 1 It appears that horses may be very comparable to humans "in vivo," since they are equally responsive to both secretagogues and have similar basal acid secretory activity.(Table 3) "In vitro" studies of isolated parietal cells have demonstrated species differences in sensitivity to specific secretagogues. Canine parietal cells can be stimulated by carbachol, gastrin, or histamine, with carbachol being the most potent. Conversely, human and rat parietal cells are strongly stimulated by histamine and only weakly stimulated by gastrin and carbachol. The horse appears to be more like the human and rat, since histamine is the most effective secretagogue of isolated equine parietal cells, followed by gastrin and carbachol.34
Acid and Sodium Concentrations Histamine versus Pentagastrin
PARAMETER EQUINE HUMAN EQUINE
Pentagastrin PG or HI** Histamine
Basal Max. Basal Max. Basal Max.
Stim. Stim. Stim.
[HI] mEq/L 21-45 35-57 10-40 70-120 28-45 75 120 [Na ]mEq/L 45-92 75-146 30-70 15-30 62-123 23-70
* from ref no.66
** from ref no.42
In this study, mean acid concentration during maximal histamine stimulation was markedly greater than that observed during maximal pentagastrin stimulation. Maximal acid concentrations in other species have ranged from 100-140 mEq/L17,39, irrespective of stimulant. Individual maximal acid concentrations in these horses ranged from 75-125 mEq/L during histamine infusion. In contrast, in pentagastrin stimulated horses, the maximal acid concentration rarely reached 75 mEq/L, which was reported44 to be a major difference between horses and other species. Apparently,a nonparietal sodium rich fluid component of gastric secretion was strongly stimulated by pentagastrin.
Sodium outputs increased coincident with acid outputs
during pentagastrin stimulation in horseS44; whereas, in the
studies described here, sodium concentration decreased as
acid concentrations increased during histamine stimulation,
and the sodium output was constant as the acid output
increased to a maximal level. (fig.2) Thus, in the horse as
in other species, histamine appears to stimulate a purely
parietal secretion, with acid concentrations rising to
expected levels and a reciprocal decrease in sodium
150 00 0 O
~100- 0~ 0t0 A;
AO AA A
0 o AA A AA HI
2 50 A A
I I I I I I
0 30 60 90 120 150
ACID OUTPUT (pEq/kg/15min)
Figure 2. Regression analysis of acid vs. sodium outputs during pentagastrin (0) and histamine (A) infusion.
The exact origin of the histamine which stimulates
parietal cells and the mechanisms involved in its release have not been clearly documented in most species24, but it is known to act via H-2 receptors on parietal cells as a paracrine mediator.2,20 In rats, enterochromaffinlike (ECL) cells are the major histamine source in the gastric mucosa and gastrin has been shown to stimulate its release from them. 22,24 Rabbit and human gastric mucosal preparations also release histamine in response to gastrin but the cells responsible have not been identified.22 Few ECL cells are found in the normal human and canine mucosa; however, many mast cells are observed and may serve as the histamine source.24 Human patients with hypergastrinemia appear to have increased numbers of ECL cells in their gastric mucosa.24 The density of ECL cells in equine gastric mucosa is greatest in the pyloric gland region.70 Fewer ECL cells have been identified in the fundic gland mucosa and in the proximal duodenum.
In this study, we determined that histamine can be used as a stimulant of gastric acid secretion in horses. Histamine stimulated mean maximal acid output was comparable to that previously reported with pentagastrin.44
In contrast, the mean maximal acid concentration in response to histamine was much greater than that observed during maximal pentagastrin stimulation. These findings suggest that in horses, histamine stimulates purely parietal secretion, while pentagastrin stimulates the production of gastric contents that appear to be both parietal and nonparietal in origin.
THE EFFECT OF PYRILAMINE MALEATE PRETREATMENT ON PENTAGASTRIN-STIMULATED EQUINE GASTRIC SECRETION
The interaction between gastrin and histamine
receptors has long been controversial.1.3 The ability of histamine receptor antagonists to eliminate pentagastrinstimulated gastric acid secretion supports the indirect action of gastrin on parietal cells.21,3137,48,49,51.53,54 This inhibition of gastric secretion has been limited specifically to histamine H-2 receptor antagonists. The administration of antihistamines directed at H-i receptors has not been shown to suppress acid production during stimulation with various secretagogues, including pentagastrin. 11,29,71 Although differences in the relative sensitivity to various secretagogues have been observed between species, the characteristics of gastric contents during maximal stimulation with histamine or pentagastrin are similar, with the exception of the horse.
Equine gastric contents during maximal stimulation with histamine have a high concentration of acid and low concentration of sodium as is characteristic of parietal secretion in other species, (see Chap. 2) whereas, pentagastrin-stimulated equine gastric contents have a high concentration of sodium and a relatively low acidity.44 Histamine stimulation of gastric secretion induces a maximal acid outputs (MAO) equal to those of pentagastrin-stimulated horses. (See Chap. 2) In horses, as in other species, it is necessary to administer an H-1 antagonist prior to histamine infusion to prevent side-effects.1,29,65 During gastric secretory studies on horses stimulated with histamine, pretreatment with the H-1 antagonist, pyrilamine maleate, resulted in a short-lived decrease in basal volume, acid concentration and acid output. (See Chap.2) The effect of this pretreatment on pentagastrin-stimulated gastric secretion is thus far unknown. Since we planned to further investigate the species specific disparity in the histamine and pentagastrin-stimulated secretory responses, we designed this study to consider the effect of pyrilamine maleate pretreatment on pentagastrin-stimulated gastric secretion.
Materials and Methods
Two mixed breed, one Thoroughbred and one Arabian [2 mares, 2 geldings] ranging from 3 to 20 years were used in this study. The horses were all healthy and ranged in weight from 370 to 490 kg. They were maintained on grass pasture and provided coastal hay ad lib and 12% protein grain twice daily. All horses were deformed every 2 months, vaccinated for encephalomyelitis and tetanus every 6 months and were free of clinical signs of gastrointestinal disease. Chronic indwelling gastric cannulas as described by Campbell-Thompson and Merri tt3l had been prepared in these horses 1 to 24 months prior to this study. All studies were approved by the University of Florida IACUC.
Horses were fasted with free choice water for 20 hours prior to an experiment. Experiments were performed with no less than a one week interval between them. The horses were loosely restrained in the laboratory for the entire
experiment. The gastric cannula was opened and allowed to drain by gravity for 15 to 20 minutes, while a indwelling jugular catheter was emplaced. The gastric contents were collected into IL fluid bags suspended from a surcingle. Experimental protocol
Gastric contents were collected in 15 minute aliquots and were filtered through gauze prior to analysis. Collections were measured for volume and if available, a 50 ml sample was saved for further analyses. Analyses were performed either immediately or samples were frozen for later analysis. Each experiment lasted 3 hours. During the first 45 minutes [basal collection], no treatment was given. In the treatment experiments, pyrilamine maleate (Histavet-P, Schering-Plough NJ) was infused IV at 1 mg/kg over a 15 minute period, beginning at time t=45min. No additional treatments were given for 30 minutes. In the no treatment experiments, basal collection was continued during time t=45-90 min. At time t=90min, an IV infusion of pentagastrin [6 ptg/kg-hr] was administered by infusion pump
(Harvard Apparatus, South Natick MA) in all experiments and continued for the remainder of the experiment.
Each infusion was given in a volume of 60 mi/hr. Horses were weighed the morning of the experiment to determine the amount of pentagastrin needed. Pentagastrin was prepared by dissolving with 0.8 ml of DMSO and 90 mls of
0.9!k NaCl and filtered through a 0.22 grn cellulose nitrate filter (Corning, Corning NY) in preparation for infusion. Sample analysis
Volume was measured in a graduated cylinder. Aliquots of gastric contents were analyzed immediately for hydrogen ion concentration, using a automatic titrator (Radiometer, Copenhagen Denmark). Hydrogen ion concentration of each aliquot was measured in duplicate by titration with 0.1N NaOH to an endpoint of pH of 7.4. Output was calculated on a per kg body weight basis from the volume and hydrogen ion concentration and was expressed as gEq/kg/15 min.
Analysis of data
Statistical analysis was performed on volume, acid
concentration, and acid output data. The last 30 minutes of basal (t=30&45 min), post-pyrilamine (t=75&90 min), and two infusion-related collections (early: t=105,120&135 min; late: t=150,165&180 mmn) were compared between the studies with and without pyrilamine pretreatment. Since the same horses were used in each study, a paired t-test was performed for each of the four time periods. The time periods were not compared to each other. A p< 0.05 was considered significant.
In both studies, the volume and acid output rapidly
increased after pentagastrin infusion began. The horses had no adverse reactions during or following the administration of pyrilamine maleate. Physical characteristics of the gastric collections from both studies were the same. The mean and SEMs for each time period of both studies are shown in Table 4.
The volume, acid concentration, and acid output during the basal time period of both studies did not differ significantly.
In the pyrilamine study, the volume collected during the 30 minutes following the administration of pyrilamine was significantly less (p=0.0145) than when no pyrilamine was given. The acid concentration did not differ between studies. The acid output was significantly less (p=0.0008) in the pyrilamine study than when no pyrilamine was given. Early Infusion Collections
The volume and acid concentration did not differ
significantly between the studies, however, the acid output was significantly less (p=0.0274) in the pyrilamine studies than when no pyrilamine was given.
Mean and SEM Data from Gastric Contents Collected Before and After Pentagastrin Infusion with and without Pretreatment with Pyrilamine Maleate.
PYRIL- BASAL POST- EARLY LATE
AMINE PYRIL- INFUSION INFUSION
VOLUME NO 253.8 + 381.7 + 585.8 + 622.5 +
37.2 29.9 74.9 43.7
YES 287.5 + 181.7 + 415 + 589.2 +
27.95 34.8 102.2 71.0
[H+] NO 31.14 + 39.7 + 31.2 + 48.0 +
9.54 11.0 5.5 5.6
YES 39.5 + 31.7 + 36.4 + 51.6 +
6.1 7.0 9.0 8.2
ACID NO 18.1 + 34.1 + 46.4 + 76.8 +
OUTPUT 5.2 7.5 5.8 5.6
(iEq/kg-15 YES 24.5 + 12.4 + 26.7 + 56.5 +
mi) 3.9 2.5* 6.0 4.0 *
First half of Pentagastrin Infusion (t=105,120,&135 min) * Last half of Pentagastrin Infusion (t=150,165,&180 min)
Significantly Less than during No Pyrilamine Study
Late Infusion Collections
As during early infusions, the volume and acid
concentration were not significantly different between the
two studies. Acid output was significantly greater
(p=0.0031) in the study with no pyrilamine pretreatment.
Pentagastrin infusion resulted in secretion of gastric acid in both studies; however, the maximal acid output was affected by the administration of pyrilamine maleate. This finding was unexpected since pyrilamine pretreatment has been used as the pretreatment of choice when performing
histamine stimulation in other species;" where the use of an H-1 specific receptor antagonist does not affect the secretion of gastric acid during stimulation. In part, the confusion over whether or not histamine acts directly on gastric mucosa resulted from early classical studies in which antihistamines were found to-have no inhibitory effect
on histamine induced gastric secretion. Better understanding of histamine receptor classes has helped to clarify the role of histamine in gastric secretion and explain these earlier findings, 11,12 namely that parietal cells have H-2 receptors and histamine stimulated gastric acid secretion appears to be a H-2 specific response. 1,2,3,68,69
Histamine receptors are involved in other aspects of gastric function as well as acid secretion. Three specific receptor types, H-1, H-2, and H-3 have, to date, been
recognized. As well as regulation of acid secretion, these receptors also play a role in the regulation of gastric microcirculation and motility."' Mucosal circulation is affected by histamine primarily via H-i receptors, although H-2 and H-3 receptors may be involved. In the rat, for instance, it appears that both H-i and H-2 receptors are involved in histamine related vasodilation.72 Vasoconstriction occurs with H-I receptor stimulation in the rabbit.2'73 Pyrilamine has been shown to competitively inhibit histamine related vasodilation in the guinea pig.73 The changes, if any, in equine gastric circulation in response to histamine or histamine blocking agents are not known.
The importance of gastric mucosal circulation in the horse may be most clearly demonstrated by the relative sensitivity of the horse to gastric ulceration due to NSAID's, though no specific studies have quantified equine gastric mucosal blood flow or the regulation of flow. The administration of NSAID's results in gastric ulceration by inhibition of prostaglandins involved in mucosal blood flow and cytoprotection.47 Gastrointestinal injury can be seen endoscopically in horses with or without accompanying
clinical signs within a few days of initiation of administration of even the recommended dose of phenylbutazone or flunixin meglumine.47 Gastric microcirculation is a key component of gastric mucosal protection2 and can be rate-limiting for gastric secretion."
Secretagogues involved in stimulating gastric acid secretion result in increases in blood flow as well. Histamine, gastrin, and cholinergic agents produce vasodilator activity associated with the increasing secretory rate.7" Blood flow has been shown to be ratelimiting at high levels of stimulation and agents which decrease blood flow will also inhibit acid secretion.75 'A histamine H-2 antagonist, such cimetidine, has been shown to decrease blood flow and acid secretion during pentagastrin stimulation in cats, however, mucosal blood flow was not decreased under basal conditions.3772 Although secretion is increased by pentagastrin stimulation in this study, the secretory response may be restricted by limitations on blood flow.
Histamine H-i receptor antagonists may affect gastric microcirculation by action at histamine receptors or by
atropine-like activity at muscarinic receptors'' involved in vasodilation. In dogs, the cholinolytic properties of pipolphen, a H-1 blocker, have been shown to suppress gastric secretions.77 Agonist to H-3 receptors78-80 and antagonist of the H-2 receptor",12,31,78 result in decreased acid secretion. Pyrilamine has been utilized in numerous secretion studies and in most species does not elicit this antisecretory response.81 8283 In other equine studies pyrilamine maleate administration was followed by a decrease in basal secretion, but this decrease was not always significant.(See Chap. 4 & 5)
Histamine H-3 receptors are found in several locations including the brain, perivascular nerve terminals, pulmonary airways, and ECL cells in the gastric mucosa.3'84 These receptors appear to be involved in the autoregulation of histamine release from ECL cells during the stimulation of gastric acid secretion. ,243679,80 They may also be involved in the release of somatostatin which inhibits gastric acid secretion." Certain drugs have demonstrated simultaneous H-1 agonist/H-3 antagonist properties,85 however, it appears that pyrilamine has a low affinity for H-2 and H-3
receptors.86 The distribution of H-1, H-2, and H-3
receptors in the horse has not been documented.
j Without Pyrilamine
S* With Pyrilamine
0- I I
Basal Pretreatment Early Late Time Block
Figure 3 Mean pentagastrin stimulated acid output with and without pyrilamine pretreatment. "Basal" refers to acid output prior to any treatment.(t=30&45 min) "Pretreatment" indicates output after pretreatment with pyrilamine or from comparable period with no treatment.(t=75&90 min) "Early" represents the first half of pentagastrin infusion. (t=105,120&135 min) Output during the last half of pentagastrin infusion is labeled "late".(t=150,165&180 min)
In this study, maximal pentagastrin-stimulated acid
output was significantly decreased in horses that received
pyrilamine prior to stimulation. This finding is contrary to the results in other species where H-i receptor antagonist do not inhibit gastric acid secretion. Because the precise mechanism for this action is not clear, it is apparent that this may be yet another important equine specific finding that warrants further investigation.
Therefore, equine gastric secretion in response to pentagastrin differs not only in composition from the classical parietal response of other species and the histamine response in the horse (See Chap.2), but also by the ability of pyrilamine to inhibit a maximal acid response.
PLACEMENT OF A CATHETER FOR COLLECTION OF DUODENAL CONTENTS IN HORSES WITH CHRONIC GASTRIC CANNULAS AND ITS EFFECT ON GASTRIC SECRETION
The contents of various parts of the gastrointestinal system differ in chemical and physical characteristics between location and species. Much of the normal gastrointestinal function of the individual species relates to the anatomy of their alimentary system and the type of diet which they consume. Increasing our understanding the normal function is often complicated by those same differences. Equine gastric secretion and the composition of the contents under various conditions have been studied by numerous investigators utilizing several techniques .31,44,48-54 The development of a chronic gastric cannula 31 for collection of gastric contents has added greatly to the knowledge regarding equine gastric physiology.
Study of the function of the equine proximal duodenum, biliary system, and pancreas has been even more difficult due to an anatomical location which strictly limits surgical exposure."8,8" Up to now, attempts to study the proximal duodenum and associated structures have met with serious difficulties and been basically unsuccessful in providing additional information on the physiology of this region.
The ref lux of duodenal contents into the empty equine stomach has been observed during endoscopic gastric examination even when there is no evidence of underlying pathology.051 The importance, frequency and volume of reflux have not been evaluated. Nor has the composition of the duodenal contents which are mixing with the gastric' contents been fully analyzed. Ref lux of large volumes of small intestinal contents into the stomach is reported during various disease conditions such as anterior enteritis and small intestinal obstruction. 8990
The presence of chronic indwelling gastric cannulas in research horses has made access to the duodenum much easier for investigator and animal. This study describes how, with the aid of an endoscope passed up through a gastric cannula, a catheter can be placed into the equine duodenum for the
collection of fluid. The two objectives of this study were to determine the composition of the duodenal fluid and whether the presence of the catheter passing into the duodenum had an effect on the gastric contents collected from the cannula before and during stimulation of secretion with pentagastrin.
Materials and Methods
Five Thoroughbreds(TB), two mixed breed (MB) horses, and one Arabian (AR) [3 mares, 5 geldings] between 3 and 20 years in age, were used in these studies. (TABLE 5) For TABLE 5.
Horses used in Gastric Collection Studies.
ID AGE BREED SEX B 5 y ARAB G
D 7 Y TB G
E 20 Y TB M
H 5 Y TB G
I 6 Y TB M
J 3 Y MIXED G M 7 Y MIXED M T 5 Y TB G
both catheter studies, six of the horses (5TB,1AR) were selected. The no catheter/pentagastrin study was performed on four (2MB,1AR,1TB) horses. Two horses (1TB,1AR) were involved in both studies.(TABLE 6) The horses were all healthy and ranged in weight from 430 to 510 kg. They were maintained on grass pasture and provided coastal hay ad lib and 12% protein grain twice daily. All horses were dewormed every 2 months, vaccinated for encephalomyelitis/tetanus every 6 months and were free of clinical signs of gastrointestinal disease. A chronic indwelling gastric cannula as described by Campbell-Thompson and Merritt3 1 had TABLE 6.
Descriptions and Abbreviations for Each Study.
EXPERIMENT INFUSION CATHETER HORSES USED
ABBREVIATION (secretagogue) (via cannula) PG/NOC Pentagastrin None B,E,J,M
PG/CATH Pentagastrin Intraduodenal B,D,E,H,I,T
been prepared in these horses 1 to 24 months prior to this study. All studies were approved by the University of Florida IACUC.
Experiments were performed with no less than a one week interval between them. The horses were fasted with free choice water for 20 hours prior to an experiment. They were loosely restrained in the laboratory for the entire experiment. The gastric cannula was opened and allowed to drain by gravity for 15 to 20 minutes while a indwelling jugular catheter was emplaced. For the catheter studies, a video endoscope (WelchAllyn, Skaneateles Fall, NY) was inserted through the gastric cannula into the stomach and steered into the duodenum to a point approximately 30 cm past the area of the duodenal diverticulum. A 4.5m long stylet was threaded through the biopsy port of the endoscope until it was seen passing from the distal end of the scope. The endoscope was slowly withdrawn as the stylet was threaded through the biopsy port, watching that it stayed in place as the scope was removed. The distance from the pylorus to the end of the gastric cannula was noted. The presence of the stylet passing through the pylorus was visually confirmed prior to removing the endoscope from the cannula. The stylet was marked at the end of the cannula, after which a specially modified stallion urinary catheter
(see below) was passed over it. The mark on the stylet was used to assure that it remained in position as the catheter was introduced through the cannula into the stomach, then through the pylorus and into the duodenum. The stylet was left in place until catheter placement was believed to be complete. The completed setup (fig. 4) resulted in the
Figure 4 Cross sectional view of duodenal catheter passing through the gastric cannula into the stomach and entering the proximal duodenum.
ability to collect gastric contents from the cannula and duodenal contents from the catheter into separate containers. The fluids were collected into IL fluid bags suspended from a surcingle. The cannula and catheter were both allowed to drain by gravity for 15 minutes after completion of catheter placement. In the studies where there was no intraduodenal catheter (NOC) in place, gastric contents only were collected by gravity into IL fluid bags.
The specialized catheter was fashioned from a 137 cm stallion urinary catheter(Jorgensen Laboratories, Inc., Loveland, Co).(fig.5) It's end was opened to allow free passage of the stylet and several additional 5mm side holes were placed in the 15cm closest to the tip. It was distinctly marked at 20 cm from the tip and this mark was considered the "0cm" mark. Additional marks were made every 5cm markings from that point to the end. To fix the catheter in position, a 10cm section of silastic tubing [16mm I.D.] was placed over a 5cm section of the barrel of a 12ml syringe (Monoject, St. Louis, MO.) and the catheter
passed tightly through the wall of the silastic tubing into the lumen and out through the syringe barrel lumen.
Figure 5 Specialized catheter for collection of duodenal contents. The bold mark is signified by and is located 20 cm from tip. Multiple fenestrations within that 20 cm region of the catheter.
One end of the syringe barrel locked into the end of the silastic gastric cannula and the other was attached to the short segment of silastic tubing which locked the catheter in a fixed position. The long stylet was made by joining together 3 stylets provided with stallion catheters. The
catheter was passed until the bold mark [0cm] was 5cm aborad to the pylorus.
Gastric (in all studies) and duodenal contents (only in the "catheter" studies [CATHI) were collected in 15 minute aliquots. Gastric samples were filtered through gauze prior to analysis. Volume of both gastric and duodenal collections was measured and, if available, a 50 ml sample was saved for sample analyses. Analyses were performed either immediately or samples were frozen for later analysis. Each experiment lasted 3 hours. During the first 45 minutes [basal collection], no treatment was given. Beginning at time t=45min, pyrilamine maleate (Histavet-P, Schering-Plough NJ) was infused IV at 1 mg/kg over a 15 minute period. No additional treatments were given for 30 minutes. At time t=90min, an IV infusion of pentagastrin [6 pg/kg-hr] was administered by infusion pump (Harvard Apparatus, South Natick MA) and continued for the remainder of the experiment. At the conclusion of the catheter
experiments, the catheter position was rechecked before the catheter was withdrawn, and the cannula closed.
Each infusion was given in a volume of 60 mis/hr.
Crystalline pentagastrin (Sigma Chemical Co., St. Louis MO) was prepared by dissolving with 0.8 ml of DMSO and 90 mis of
0.9% NaCi before filtering through a 0.22 I.In cellulose nitrate filter (Corning, Corning NY) in preparation f or infusion.
Gastric and duodenal sample volume was measured in a graduated cylinder and then an aliquot was analyzed immediately for chloride ion concentration. Gastric samples were also analyzed immediately for hydrogen ion concentration. Duodenal samples were kept in an ice bath until they were analyzed for bicarbonate ion concentration. Both gastric and duodenal samples were frozen for later measurement of sodium and potassium ion concentration.
Using a automatic titrator (Radiometer, Copenhagen Denmark), hydrogen ion concentration was measured in duplicate by titration with 0.1N NaOH to an endpoint of 7.4.
Chloride ion concentration was measured in duplicate by a digital chloridometer (Buchler Instruments Div., Nuclear Chicago, Fort Lee NJ) with acid reagent (Labconco, Kansas City, Mo). Chloride standard (Labconco, Kansas City, Mo) was used to calibrate the machine prior to each experiment and after every 20 tests. Bicarbonate ion concentration was determined by back-titration method of Isenberg et al.91 with the automatic titrator (Radiometer, Copenhagen Denmark) using 0.1N NaOH to titrate to an endpoint of 8.4. Sample solutions were gassed with nitrogen washed in barium hydroxide to remove carbon dioxide prior to and during the titration and were analyzed in triplicate. Standard solutions prepared in laboratory were measured in quadruplicate prior to each experiment. Sodium and potassium ion concentration were measured by flame photometry (Instrumentation Laboratories Inc., Lexington MA) on samples which had been frozen at -200C. Samples were thawed to room temperature, and diluted in an internal standard lithium solution (Dilumat, Fisher Scientific). The machine was calibrated with known [Na ]/[K ] standards (Instrumentation Laboratories Inc., Lexington MA) prior to any analyses and after every 5 samples.
Analysis of data
For the purpose of this paper, duodenal contents were analyzed to determine electrolyte concentration ranges, but these data were not statistically evaluated. Statistical analysis was performed on volume, acid concentration, acid output, sodium concentration, and sodium output from the gastric samples. Output was calculated on a per kg body weight basis from the volume and hydrogen or sodium ion concentration and was expressed as pEq/kg-15min. The last 30 minutes of basal collections, post-pyrilamine collections, and the first and last 30 minutes of infusionrelated collections were compared between studies. A
two-way ANOVA was performed for factors of time (basal, post-pyrilamine, early infusion, and late infusion), and duodenal catheter (yes or no) and interactions of these factors. A p<0O.05 was considered significant and all pairwise multiple comparisons by Tukey test were performed.
Volume. (Table 7) The volume of gastric collection
differed significantly (p<0.01) between the designated time blocks. The volume was significantly greater (p<0.05) during the late infusion block than during all other blocks. The early infusion volume was significantly greater (p<0.05) than the post-pyrilamine volume, but did not differ significantly from the basal volumes. The basal and postpyrilamine volumes did not differ significantly. The volumes collected during the CATH study were significantly (p<0.01) greater than during the NOC study. There was no significant interaction between time blocks and catheter status.
Acid Concentration. (Table 7) Acid concentration varied significantly (p=0.001) among the time blocks with a significantly higher (p<0.05) concentration during late infusion. The [H+] was not significantly different between basal, post-pyrilamine and early infusion blocks. The NOC study had a significantly (p=0.016) higher concentration of
acid than the CATH study. No significant interactions were
found between time blocks and catheter.
Data from Study With and Without Duodenal Catheter During Pentagastrin Stimulation (Mean and SEM's)
STUDY BASAL PYRILAMINE EARLY LATE
VOLUME NOC 287.5 + 171.3 + 405.0 + 568.8 +
(ml/15 28.0 27.0 125.4 83.4 *
CATH* 541.7 + 396.6 + 653.3 + 885.8 + 29.7 32.1 66.0 26.7 *
[H] NOC** 39.5 + 32.3 + 5.2 31.9 + 52.4 +
(mEq/L) 6.1 10.0 9.4 *
CATH 26.5 + 28.9 + 3.3 20.4 + 42.4 +
4.1 3.7 2.2 *
ACID NOC 24.5 + 12.1 + 1.9 22.5 + 57.4 +
OUTPUT 3.9 6.5 3.8 *
(pEq/kg15rn CATH* 31.9 + 24.9 + 3.7 30.2 + 80.2 +
15m) 6.2 6.8 4.9 *
[Na] NOC 106.1 + 110.8 + 108.3 + 93.4 +
(mEq/L) 6.1 5.7 10.0 8.2
CATH 106.8 + 104.4 + 107.4 + 88.2 + 6.0 3.3 4.2 3.3
NA NOC 72.5 + 44.8 + 8.5 107.3 + 133.8 +
OUTPUT 10.0 40.0 29.2 *
(pEq/kg(Eq/kg- CATH* 122.1 + 88.0 + 8.0 146.4 + 167.5 + 15min)
15m) 8.4 13.6 9.9 *
* Significantly greater than NOC
** Significantly greater than CATH
* Significantly greater than all other time blocks Significantly greater than post-pyrilamine time block
* Significantly less than all other time blocks
* Significantly greater than basal & post-pyrilamine time
Acid Output. (Table 7) The output of acid decreased after the pyrilamine infusion and increased during pentagastrin infusion. These changes in acid output over time differed significantly (p<0.001). Acid output (AO) was significantly greater (p<0.05) during late infusion than during all other periods, however, the differences between outputs during each of the other times did not differ significantly. The AO was significantly greater (p=0.001) during the CATH study than the NOC study.
Sodium Concentration. (Table 7) The [Na ] was
relatively constant until the late infusion block during which time the concentration decreased significantly (p=0.01). There was no significant difference in the concentration between CATH and NOC studies.
Sodium Output.(Table 7) There was a significant (p<0.001) effect of time on sodium output. The output during late infusion was significantly greater (p<0.05) than during basal or post-pyrilamine periods, but not significantly greater than during the early infusion period. Early infusion sodium output was also significantly greater than post-pyrilamine, though it was not significantly
greater than the basal output. The output during the CATH study was significantly greater (p=0.001) than during the NOC study.
Duodenal fluid was generally thick and mucoid and tended to be dark yellow to green, in color. The bicarbonate ion concentration ranged between 19-40 mEq/L; TABLE 8.
Electrolyte Composition of Fluid from Duodenal Catheter During Basal, Post-Pyrilamine, and Pentagastrin Infusion
Collection Periods. [Mean + SEMI
DUODENAL BASAL PYRILAMINE _INFUSION
[HCO3] 23.2 + 2.6 16.2 + 1.5 33.0 + 2.3
[Na'] 147.5 + 3.9 151.5 + 6.7 144.9 + 2.4
[K+] 4.5 + 0.2 4.4 + 0.3 3.6 + 0.1
[Cl-] 118.1 + 5.5 119.6 + 5.2 115.1 + 6.1
it decreased noticeably in the post-pyrilamine collections and was increased during pentagastrin infusion. The fluid had a high concentration of sodium (120-190 mEq/L) and low concentration of potassium (2.4-5.5 mEq/L) compared to that of gastric contents, in which the [Na'] ranged between 25-
125 mEq/L and [K+] between 6-20 mEq/L. Chloride ion concentrations were in the range of 95-160 mEq/L, whereas the gastric contents had [Cl]I ranging from 130 to 180 mEq/L. The mean concentrations of sodium, potassium, and chloride were consistent throughout the experiment.
The placement of a duodenal catheter through the
gastric cannula required variable amounts of manipulation of the videoendoscope. The time required for introduction of the video endoscope into the duodenum and passage of the stylet and threading of the catheter ranged from 10 to 2+5 minutes. The differences related in position of the pylorus relative to the cannula and the dexterity of the investigator on that day. The horses did not appear to be bothered by the process nor by the presence of the catheter during the experiments. Catheter experiments proceeded as the no catheter experiments did. Gastric contents were easily collected from the cannula even with the catheter in place. During some time periods, no fluid was collected from the duodenal catheter. However, this did not indicate
blockage as fluid was collected during the next time period as patency was checked injection of air through the catheter.
In the catheter experiments, gastric contents had
significantly greater volume, acid output, and sodium output
as well as a significantly lower [Hi1 The presence of the catheter passing through the pylorus may have allowed additional fluid ref lux from the duodenum into the stomach. This may have occurred due to capillary action along the catheter or by preventing complete closure of the pylorus. The duodenal fluid had a high concentration of sodium and reflux: of this fluid may account for increased sodium output and volume of gastric contents in CATH experiments. The' acid response to pentagastrin with the duodenal catheter was similar to previous studies in the horse ;31,'44 however, during pentagastrin stimulation, the gastric contents [H+1. in the CATH study was less than in the NOC study. We suggest that this may have been due to dilution of the gastric secretions by the fluid refluxing from the duodenum around the catheter.
The increased acid output observed in the CATH study was not related to the ref lux of duodenal fluid. The post-
pyrilamine period decrease in acid output in the CATH study was not as dramatic in the NOC study. The increasing acid output during the early and late infusion periods of the CATH were closer to the normal pentagastrin response observed in horses without pyrilamine pretreatment(See Chap.3) than to an unusually profound response to pyrilamine in the NOC study. Decreased maximal acid output in horses stimulated with pentagastrin following the administration of pyrilamine maleate may be due to decreased mucosal blood flow. (See Chap. 3) The placement of the duodenal catheter may also potentiate gastric acid secretion by mechanical stimulation2,3811,18 in the gastric antrum resulting in the local release of gastrin or acetylcholine, thereby lessening the effect of the pyrilamine. In humans and dogs, distention of the antral region has been shown to augment histamine or gastrin- related secretion of acid.2'3"8 Although the catheter was not large and did not distend the antrum, it did contact the gastric mucosa and may have stimulated local intramural reflexes23"17,18 involved in the secretory response.
Sodium output increased dramatically during
pentagastrin infusion in both the CATH and NOC studies,
although, the output was significantly greater in the CATH studies. The infusion related increase in sodium output was consistent with the apparently equine specific response to pentagastrin. 11,44,51 Monitoring of sodium ions in the stomach has been used to assess duodenogastric ref lux in humans. 9293 The sodium rich fluid collected from the equine gastric cannulas is probably of duodenal origin5'5' and the ref lux of this sodium rich duodenal fluid around the duodenal catheter could explain the increased sodium output during the CATH study.
The equine gastric cannula model has been beneficial in the understanding of equine gastric physiology and the development pharmaceuticals for the treatment of gastric ulcers. 3149,53'54 It appears that this model may also allow further investigation of equine gastric and small intestinal physiology. The duodenal contents had a high concentration of sodium and chloride and low concentration of potassium and is most likely the fluid which dilutes parietal secretions during gastric collections. The passage of a duodenal catheter and collection of duodenal contents does not prevent normal acid stimulation in response to pentagastrin, but may enhance ref lux of duodenal contents into the stomach.