Gastric acid and pepsin secretion in the conscious young horse


Material Information

Gastric acid and pepsin secretion in the conscious young horse
Physical Description:
xi, 131 leaves : ill. ; 29 cm.
Campbell-Thompson, Martha L., 1954-
Publication Date:


Subjects / Keywords:
Gastric Acid -- secretion   ( mesh )
Pepsin A -- secretion   ( mesh )
Horses -- physiology   ( mesh )
Veterinary Medicine thesis Ph.D   ( mesh )
Dissertations, Academic -- Veterinary Medicine -- UF   ( mesh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1988.
Includes bibliographical references (leaves 118-130).
Additional Physical Form:
Also available online.
Statement of Responsibility:
by Martha L. Campbell-Thompson.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 022021265
oclc - 22441575
System ID:

Full Text







Copyright 1988


Martha L. Campbell-Thompson

To Floyd and my parents


I would like to acknowledge my entire committee for

creating a challenging graduate program, and specifically, my

mentor, Dr. Al Merritt, who made the entire program possible

by accepting me as his graduate student. Al guided me through

these last years, always encouraging self-taught discoveries.

I thoroughly enjoyed working with someone who was always

available for help or criticism, but was never over-powering

with his own opinions. I would also like to thank Dr. George

Gerencser who allowed me to write a paper for his course

which created the basis for my understanding the processes of

acid secretion and which continues to serve me to this day.

Dr. Stephen Russell was especially kind to remain active on

the committee despite moving to an endowed professorship at

the University of Kansas Medical Center. Finally, I am

indebted to Dr. James McGuigan who introduced me to his lab

and the incredible possibilities for research with the

gastric parietal cell. I learned to isolate viable canine

parietal cells and was thrilled to discover "life after


I wish to acknowledge my husband, Floyd, for his never-

ending support, guidance and encouragement. I used his

computer (IBM XT compatible), software (WordPerfect@,

WordPerfect Corp.; Twin@, Mosaic Software Inc.) and printer

(HP Laserjet Series II) to write this dissertation. I need to

thank Wesley Smith who introduced me to the IBM spreadsheet

and created the first template for my experiments. Drs. David

Williams and Colin Burrows were also kind to allow me to use

their computers and software. Susan Lowrey contributed

innumerable hours of technical assistance and various

students helped on this project through the years. Without

this assistance, I would not have been able to complete this

project in this amount of time.

I am privileged to be part of a family filled with

incredible loving people who always encouraged me in this

endeavor. I am thankful for the flexibility in this program

which allowed me to pursue several personal projects

including building a house and completing the American

College of Veterinary Surgery Boards. I would also like to

thank our friends, Will and Donna Wheeler-Harding, who have

shown me everything is possible if I'm willing to work

(train) for it.


ACKNOWLEDGMENTS...................................... iv

LIST OF TABLES ...................................... viii

LIST OF FIGURES...................................... ix

ABSTRACT............................................. x


I INTRODUCTION ........ ............................ 1

Methods for Studying Gastric
Acid Secretion........................... 2
Gastric Mucosal Anatomy....................... 8
Parietal Cell Function........................ 12
Phases of Gastric Acid Secretion............. 20
Pepsinogen Secretion Rate..................... 28
Gastric Fluid Composition..................... 31

II REVIEW OF LITERATURE.......................... 35

Peptic Ulceration............................. 35
Equine Gastric Secretory Physiology.......... 44

III MATERIALS AND METHODS......................... 46

Animals and Surgical Preparation............. 46
Methodology................................... 49
Experimental Design........................... 50
Statistical Analysis.......................... 55

IV RESULTS ...................................... 57

Animal Preparation and Necropsy Findings......57
Gastric Secretory Studies .....................58
Control Studies.............................. 60
Pentagastrin Stimulation Studies............. 61
H2-receptor Antagonist Studies............... 66

V DISCUSSION AND CONCLUSIONS.................... 81

Equine Gastric Fluid Secretion in
Comparison to Other Species................ 81
Future Directions ............................ 98
Summary ...................................... 99
Conclusions .................................. 102


A DATA TABLES .................................... 103

B ANOVA TABLES AND P VALUES .................... 106

C CORRELATION MATRICES ......................... 117

REFERENCES ........................................... 118

BIOGRAPHICAL SKETCH .................................. 131



Table Page

3.1 Animal signalment.............................. 47

3.2 Treatment scheme for protocol IV.............. 52

4.1 Phenol red recovery rates ..................... 59

4.2 Control gastric fluid variable averages....... 62

4.3 Maximal acid output during
step-dose pentagastrin infusion.......... 64

4.5 Statistical analysis of replicate samples..... 80

5.1 Basal and stimulated gastric fluid secretion
in various species ....................... 90

5.2 Acid Output during basal and stimulated
conditions in various species............ 91

5.3 Calculated parietal cell mass in the horse.... 93



Figure Page

4.1 Basal acid output in 13 horses................ 63

4.2 Acid output following pentagastrin infusion... 65

4.3 Hourly secretion rate for protocol IV......... 72

4.4 Hourly pH for protocol IV...................... 73

4.5 Hourly acid concentration for protocol IV..... 74

4.6 Hourly acid output for protocol IV............ 75

4.7 Hourly pepsin concentration for protocol IV... 76

4.8 Hourly pepsin output for protocol IV.......... 77

4.9 Hourly osmolality for protocol IV............. 78

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



Martha L. Campbell-Thompson

August 1988

Chairman: Alfred M. Merritt, DVM
Major Department: Veterinary Medicine

Gastric cannulation was performed in young horses to

determine the composition of parietal and nonparietal

components of gastric fluid during basal (fasting),

stimulated and inhibited conditions. Horses were fasted for

24 hours after which a 5-hour collection of gastric fluid was

made. Intravenous saline (200 ml/h) served as a control

treatment (29 experiments in 13 horses). Basal volume (290

77 ml/15-min), pH (2.0 0.5), and acid concentration (39

15 mEq/L) and output (56 30 AEq/kg/15-min) were constantly


Stimulation of the parietal component of gastric fluid was

achieved with pentagastrin, a synthetic analog of the

hormone, gastrin. A dose-response curve showed that 6

gg/kg/h, by intravenous infusion, stimulated maximal acid

output (5 experiments in 2 horses). Pentagastrin at 6

Ag/kg/h IV for 4 hours produced maximal acid output during

the second and third hours by doubling flow rate to 610 260

ml/15-min and increasing acid concentration to 55 mEq/L.

Basal acid output (210 AEq/kg/h) was approximately 30% of the

maximal acid output (630 gEq/kg/h).

The histamine H2-receptor antagonist, ranitidine (0.5

mg/kg intravenous bolus), minimally effected flow rate but

inhibited acid concentration and output (10 experiments in 5

horses, ANOVA, P<0.05). Ranitidine was also effective in

reducing pentagastrin-stimulated acid output to basal levels

by decreasing both flow rate and acid concentration.

Basal pepsin concentration was highly variable (110 34

Ag/ml; 10 experiments in 5 horses) and first increased and

then decreased, during pentagastrin infusion, suggesting a

washout of stored pepsinogen with subsequent dilution as

gastric flow rate increased (P > 0.05). Ranitidine decreased

pepsin concentration for one hour, probably as a result of

decreased luminal acid (P > 0.05).

In comparison to other species, equine gastric fluid had a

low acid concentration and output during maximal stimulation,

despite a large flow rate and was constantly hypotonic.

These data suggest that there is a prominent nonparietal

component to equine gastric fluid secretion and this species

may be useful for studying the regulation of the nonparietal



Gastric and duodenal ulceration has become a significant

disease in young and adult horses and its clinical and

pathological consequences have only recently been described

(30). While it appears that gastric ulceration is produced by

administration of nonsteroidal anti-inflammatory drugs, the

etiology of the spontaneously occurring gastroduodenal

ulceration seen only in suckling foals remains unknown. This

is not surprising since the etiology of duodenal ulceration

in people, although studied for decades, is also unknown

(103). The mechanisms regulating human gastric secretion, as

well as a variety of factors which appear important in the

development of human gastroduodenal ulceration, have been

studied extensively (103). Knowledge of equine gastric

physiology would provide an understanding of some of the

factors responsible for producing equine mucosal ulceration,

as well as a basis for its treatment (103).

Development of a technique to study equine gastric

secretion was necessary. This technique would not only

further our knowledge of gastric physiology in this species,

but would also serve as another animal model for studying

basic gastric physiology. Studies of basic gastric physiology


in different animal species have revealed variations in

mechanisms of regulating gastric secretion (40). Since the

young horse can develop spontaneous duodenal ulceration,

gastric physiology in this species is particularly

interesting as the young horse could provide an animal model

for duodenal ulceration.

This dissertation presents current knowledge concerning

the regulation and composition of equine gastric fluid based

on studies performed in normal young horses using a new

gastric cannulation technique.

Methods for Studying Gastric Secretion

Various methods are available for studying gastric

physiology (45). In the conscious person or animal,

nasogastric aspiration can be used with about an 80% recovery

rate of gastric contents (125). There can be contamination of

the fluid with saliva and duodenal contents, although people

can be instructed to not swallow their saliva and dental

suction can be employed. Even without these measures,

salivary contamination is estimated to be less than 10% of

the gastric contents.

The most common technique of studying gastric secretion in

animals is the gastric fistula, where an adhesion is created

between the stomach and body wall and through which a

collection tube (Thomas cannula) is inserted. The fistula

technique leaves the stomach with intact innervation and

exposed to the normal flow of ingesta and secretions. With


correct attention for ventral placement and complete

collection of fluid, quantitative measurements of gastric

secretions can be made. However, gastric secretions may be

contaminated with swallowed saliva or regurgitated duodenal

contents and it is difficult to measure food-stimulated

secretions by this technique. Despite these limitations, the

gastric fistula remains the most common method employed for

chronic studies in laboratory animals.

The first important information about gastric physiology

in people was collected in 1825 by a military surgeon,

William Beaumont, who carefully studied a patient, Alexis St.

Martin, who had a chronic gastric fistula following an

abdominal gunshot wound (45). Beaumont made observations over

a 4 year period and produced the first accurate description

of gastric fluid secretion during different emotional states.

His observations were later confirmed by Pavlov and others in

different species.

Much of the knowledge concerning the control of acid

secretion has been derived from experimental surgical

preparations in dogs. These have allowed the study of neural,

humoral and endocrine factors that determine the rate of

secretion (68). To determine the effect of the central

nervous system on acid secretion, uncomplicated by the

effects of food in the stomach, acid output can be easily

measured using fundic pouches. Such preparations make it

possible to sample gastric juice without food contamination


while maintaining alimentary tract continuity. Pavlov is

credited with numerous basic observations concerning

gastrointestinal physiology and developed the innervated

fundic pouch (Pavlov pouch) as a means to study un-

contaminated gastric secretions in response to food (45).

Vagally innervated pouches permit extremely sensitive

assessments of changes in gastric secretion but usually

involve some degree of mucosal denervation as well as the

consequences of diverting the normal food stream from the

pouch mucosa. Heidenhain introduced the denervated fundic

pouch (Heidenhain pouch) as a means to study the intrinsic

control of gastric secretion (45). The sympathetic nervous

supply is intact as these nerves accompany the blood vessels.

These pouches usually have a low maximal secretary flow rate

which makes small but significant changes in secretion rate

difficult to measure. Antral pouches may also be prepared.

Total gastric pouches as developed by Dragstedt and Ellis

have a high flow rate but are generally not used due to

difficulties in maintaining the animals (45).

Pavlov also introduced the technique of permanent double

esophagostomies which allows sham-feeding in a dog with a

gastric fistula (45). The dog was fed through the distal

stoma between experiments. Komarov and Marks devised a 2-

stage esophagostomy where the esophagus is everted only

during the experimental period to divert food.


The first attempt to depart from standard procedures using

a fistula or pouch involved exteriorization of the entire

stomach with the circulation intact (dog flap or vascularly

perfused mouse stomach) (45). This technique then allowed the

additional measurements of oxygen and glucose uptake, CO2

production and blood flow and volume changes in gastric

circulation to be made. The Shay rat preparation involved

ligation of the pylorus under general anesthesia (45). The

rat was allowed to recover and, usually 4 hours later, was

re-anesthetized and the entire gastric contents removed.

Mucosal function has been studied extensively in isolated

preparations by mounting a mucosal segment (usually devoid of

the muscularis) in an Ussing chamber, where the serosal

(nutritive) and mucosal secretaryy) chamber solutions can be

modified independently (45,119,138). Electrophysiological

studies could then be performed directly on the mounted

segment to determine tissue resistances or potential

differences. Consistent relationships between variables, such

as tissue secretary capacity, ionic composition of bathing

solutions, ion fluxes and electrical phenomena have been

established. Microelectrodes are also available to study

individual cells and intracellular ion and pH changes (137).

Difficulties in determining specific parietal cell

function in the intact mucosa have led investigators to

develop techniques for gland and cell separation (18,138).

For the gastric epithelium, enzymatic digestion using


collagenase or pronase produces primarily gastric glands.

Addition of calcium chelating agents such as EDTA to disrupt

tight junctions will isolate gastric cells. Advantages of

isolated glands include a more intact system with tight

junctions, normal shape and polarization and cell-cell

communications, the ability to follow pH changes and an

enrichment of parietal cells from 20% in the intact mucosa to

50% by volume. Advantages of isolated cells include similar

responses as the gland preparation with the ability to

further enrich specific cell types. Cell viability is

generally greater than 90% and the cells can also be

maintained in tissue culture.

Isolated cell preparations have allowed the determination

of parietal cell physiology previously clouded by the

presence of the numerous other cell types in the gastric

mucosa (137). For studies of amino acid uptake, ion fluxes or

membrane receptors, isolated parietal cells have been

necessary. Three techniques are used to enrich parietal cell

content: 1) the production of gastric glands automatically

enriches the parietal cell content as stated earlier; 2)

density gradient purification can be used with either rat or

canine parietal cells and yields a high purity of viable

cells; 3) cells can be separated by an elutriator rotor which

fixes the cells by centrifugal forces and by concurrent

solvent flow, elutes the cells based on their density (138).

Parietal cells are the largest cells in the gastric mucosa.


Indices of parietal cell responses to stimulation include

oxygen consumption, morphological transformation (coalescence

of cytoplasmic tubulovesicles into secretary canaliculus) and

aminopyrine uptake (18). Weak bases, such as aminopyrine,

partition between the gastric circulation and lumen as a

function of their pKa; at cytoplasmic pH, the bases are

unionized and can diffuse through lipid membranes such as

tubulovesicles and secretary canaliculi. Protonated bases are

trapped in the acid spaces and thus their uptake by the

parietal cell can be used as an index of parietal cell

responsiveness to stimulation. Aminopyrine accumulation

provides a quantitative index of the parietal cell

responsiveness to stimulation which is reflected in the

quantity of sequestered acid but not actual secretion rates

(138). Isolated cell preparations offer advantages such as

standardized experimental conditions and straight-forward

stimulus-response detection. On the other hand, all in vitro

models involve certain steps in their preparation that may

influence, or even damage, the receptors on the cell surface.

Many of the recent breakthroughs in parietal cell

physiology can be attributed to the characterization of

enzymatic activities associated with membrane fractions,

specifically the membrane vesicles from which the H+-K+-

ATPase was isolated (55,127). The highly glandular nature of

the gastric epithelium as well as the invagination of the

parietal cell apical surface makes the gastric mucosa


abundant in membrane material. Morphometric data yield values

of 400-500 cm2 of parietal apical membrane for each cm2 of

mucosa (138).

This discussion is intended to be a brief overview of 160

years of methods involved in the study of gastric secretion.

New information is available monthly on cellular- and

membrane-associated phenomena. Increasingly simplified models

facilitate the ability to explain the phenomena but on the

other hand, subtle regulatory processes present in the intact

animal may be lost and one may be misled as to the prominence

of a process in the simplified model. It is essential to

refer back to the intact animal or at least intact tissue

data to see whether or not information obtained in isolated

cells is consistent at all levels (138).

Gastric Mucosal Anatomy

The gastric mucosa is arranged in folds, or rugae, with

the surface covered by a simple columnar epithelium of

surface mucous cells (86). The glandular mucosa is divided

into 3 regions, cardiac, fundic and pyloric, based on

histologic features. The cardiac region is located adjacent

to the esophageal region at the transition between squamous

and columnar epithelium and there is marked species variation

in its proportion of the total glandular surface (73). The

cardiac glands contain mucus and undifferentiated endocrine

cells. The fundic gland region contains mucous, chief,

parietal and endocrine cells and comprises the majority of


the glandular epithelium in the human and dog (-80%) (73).

The fundic gland is divided into the following 3 regions: 1)

isthmus-- surface mucous cells; 2) neck-- parietal and mucous

cells; 3) base-- chief and parietal cells. Pyloric glands are

located in the antrum and contain primarily mucous cells and

a few endocrine cells, such as gastrin cells which produce

gastrin. One or more simple or branched tubular gland opens

into a gastric pit which is also lined by mucous cells.

Pyloric and fundic glands also have enterochromaffin cells,

which secrete serotonin, and several types of endocrine

cells, which contain somatostatin and vasoactive intestinal

peptide (86). Some of these peptides may be released into the

tissue (paracrine secretion) rather than into the

circulation. One study has shown that the equine glandular

mucosa has gastrin and somatostatin positive cells in the

antrum and proximal duodenum (91). Other cells of the mucosa

include mast cells and plasma cells (for IgA synthesis)

within the lamina propria. The muscularis mucosa is generally

2-3 cell layers followed by a stratum compactum in some


The equine gastric mucosa is composed of approximately 1/3

stratified squamous and 2/3 glandular epithelium (135). This

composition is similar to that of rodents, but distinct from

that of most other species, where the gastric mucosa is

composed entirely of glandular epithelium (73). The mucosa of

the esophageal region is demarcated from the cardiac zone by


a fold known as the margo plicatus. The portion proximal to

the margo plicatus forms a large sac known as the saccus

cecus and is lined by stratified squamous epithelium. The

squamous mucosa is devoid of glands, but has a thick layer of

keratinized epithelium which is probably the main barrier to

acid and pepsin.

Normally, the gastric epithelial cells and inter-cellular

tight junctions provide an almost completely impermeable

barrier to the back-diffusion of H+ (101,128). The geometry

of the glands makes them relatively inaccessible and

contributes to the low permeability of the luminal surface to

cations (120). The "gastric mucosal barrier" is a descriptive

term which refers to the ability of the gastric mucosal

epithelium to maintain a large lumen to mucosal H+ gradient

under physiological conditions (39). Components which have

been used to describe the functional integrity of the

barrier, in addition to very low H+ diffusion from the lumen

to the mucosa, include a low diffusion of Na+ and K+ from

mucosa to lumen and the maintenance of a lumen-negative

transmucosal potential difference. Although no anatomical

barrier has been described, it has been suggested that the

mucous gel and epithelial phospholipids are constituents


The parietal cell is responsible for acid and intrinsic

factor secretion and is oval to pyramidal in shape (86). The

most conspicuous features of the cell are intracellular


canaliculi, abundant smooth endoplasmic membranes called

tubulovesicles and abundant, large mitochondria. The

intracellular canaliculi are a structural specialization

which are open to the lumen but may become internalized in

nonsecretory states. Parietal cells have long and numerous

microvilli on their luminal surfaces and canalicular walls

which serve to increase surface area greatly. During

secretion at high rates, the microvillar surface increases

while the tubulovesicular membrane decreases. A detailed

study of these ultrastructural changes has led Forte and

coworkers to propose a mechanism of membrane formation based

on the fusion and recycling of tubulovesicles to and from the

apical membrane (56). In addition to membranes, the

morphological transformations seem to involve actin

microfilaments which have been suggested as the means by

which the microvillar membrane is withdrawn from the surface

and reconstituted into the tubulovesicular membrane (127).

The chief cell is responsible for pepsinogen secretion and

is located at the bases of glands (86). Pepsinogen is also

secreted by mucous neck cells of cardiac, fundic and pyloric

glands (72). Chief cells store pepsinogen in encapsulated

zymogen granules which are released by exocytosis (72,74).

Zymogen granules are synthesized and encapsulated in the

Golgi body. When first formed, the granules are argyrophilic

and stain poorly, but as they mature and migrate toward the

apex, they are argyrophobic and take a crystal violet stain.


As granules accumulate in the cell, a negative feedback

mechanism slows down new synthesis. When the cell is full of

granules, synthesis ceases except for a small steady resting

secretion rate which is approximately one fifth the maximal

stimulated rate. During maximal stimulation, the proenzyme is

secreted without going through the granule stage. This may

occur before an increase in water secretion is observed

resulting in an initially high concentration of pepsinogen in

gastric juice of low volume. There appears to be a delicate

balance between synthesis, storage and secretion so that the

chief cell is capable of both immediate and prolonged

continuous secretion.

Parietal Cell Function

The parietal cells, which constitute about one tenth of

the mucosal volume, increase the concentration of hydrogen

ions (H+) one million-fold in the fluid extracted from plasma

('4 x 10-8 M, pH 7.4) and secreted as gastric juice (-0.15 M)

(18,115,137). A large amount of energy is required to

transport H+ against this large concentration gradient and

the parietal cell's high mitochondrial content indicates the

important contribution of oxidative metabolism to the energy

economy of the cell.

Davenport developed a hypothesis for acid secretion which

still holds today (39). It is believed that water is

hydrolyzed to provide a H+ for secretion coupled to potassium

(K+) exchange and chloride (Cl-) secretion. The hydrolysis of


carbonic acid, which is formed by the hydration of carbon

dioxide through the action of carbonic anhydrase, neutralizes

the resultant hydroxyl ion and produces a bicarbonate ion,

which is removed by the venous blood (alkaline tide)

(39,138). Administration of pentagastrin, histamine or cyclic

AMP enhances the activity of carbonic anhydrase. Experiments

in vitro have demonstrated that the gastric carbonic

anhydrase can be separated into two isoenzymes and the

phosphorylation of one by a cAMP-dependent protein kinase

sharply increases its activity (138).

In parietal cells, two primary active ion transport pumps

have been demonstrated from isolated membrane studies (18).

The classic Na+-K+-ATPase is localized to the basolateral

membrane while the H+-K+-ATPase is unique to the microvilli

of the secretary canaliculi. In isolated cells or mucosal

segments, sodium, potassium and chloride play essential roles

in the process of acid secretion.

Biochemical studies have led to the establishment of the

H+-K+-ATPase as the gastric proton pump (57,119). This enzyme

was first identified by Forte and coworkers in the form of K+

phosphatase activity in rabbit and frog mucosa (55). The

causal relationship between the K+-ATPase and H+ was provided

in studies which demonstrated that a K+-ATPase-rich

microsomal fraction from canine mucosa would extract protons

from the medium, implying that the microsomes were inside-out

sacs of the membrane generating HCl in the intact cell (57).


Finally, Sachs et al. demonstrated that H+ accumulation was

the result of an ATP-driven electroneutral exchange of

intramicrosomal K+ by H+ (127). The precise number of H+

transported per ATP consumed is still controversial, with a

number of laboratories claiming measured H+-to-ATP ratios of

one or two. Substituted benzimidazoles, such as omeprazole,

have been identified as specific H+-K+-ATPase inhibitors


The well-known enzyme associated with Na+ transport, the

Na+-K+-ATPase, has been identified in the parietal cell

fraction from several species (119). The enzyme shows a

positive correlation with basal acid output, is stimulated by

cholinergic agents, and is inhibited by atropine, epinephrine

and cAMP (138).

The parietal cell is stimulated to secrete acid by

interactions between 3 chemical messengers, acetylcholine,

gastrin and histamine (115). Each agent utilizes different

modes of delivery to affect the parietal cell (18). Each

agent binds to a specific receptor on the parietal cell and

this union triggers a series of biochemical steps within the

parietal cell (137). A final common pathway involves the

activation of the H+-K+-ATPase with the resultant HCl

secretion (137). Gastric fluid stimulated by acetylcholine is

rich in all its components (water, acid, pepsin, mucus, ions)

(119). In contrast, gastrin and histamine stimulate mostly

water and acid, but much less pepsin and mucus (119).


Acetylcholine is released by vagal stimulation and

activates muscarinic cholinergic receptors on the parietal

cell after crossing a short synaptic gap from the

postganglionic fiber (66). In rabbit parietal cells, the

response to the cholinergic agent, carbachol, was less than

to histamine (40). In contrast, in canine cells carbachol

produced a sustained increased response which was greater

than that produced by histamine.

Acetylcholine stimulates parietal cell acid secretion by

increasing intracellular calcium (115,128,137). A calmodulin-

regulated, ATP-dependent calcium pump that could maintain

active calcium extrusion was found in the basolateral

membrane of parietal cells (128). Calmodulin regulates the

activity of several enzymes such as adenylate cyclase,

phosphodiesterase, phospholipase A2 and protein kinase and

membrane phosphorylation, and could thus regulate both

synthesis and breakdown of cAMP and synthesis of a protein

kinase or prostaglandin.

Gastrin is present in cytoplasmic secretary granules in

gastrin cells, which are interspersed singly or in small

clusters among the other epithelial cells in the antrum

(103). The major tissue form is the heptadecapeptide (G-17)

(103). Gastrin release appears to be controlled by a number

of factors, including direct action of luminal agents on the

gastrin cell and both stimulatory and inhibitory cholinergic

effects (40,65,66). Gastrin produces a small but definite


increase in parietal cell response in canine cells which is

markedly potentiated by histamine; however, gastrin is

relatively ineffective in the rabbit (32,138). Gastrin has

been linked to enhanced calcium influx or intracellular

mobilization but solid data is lacking.

Gastrin release is inhibited by acid bathing the antral

mucosa, probably by direct action on the G-cell, with marked

suppression at pH 2 and complete inhibition at pH 1 (36,66).

Gastric acid inhibition by antral acidification may also be

mediated by prostaglandins and somatostatin, both acting as

paracrine regulators of gastrin release (16,57,133). There is

also evidence that vagal fibers may inhibit gastrin release


Histamine, whose release mechanism is unknown, is stored

by mast cells in the human and canine mucosa and by an

enterochromaffin-like endocrine cell in the rat mucosa

(86,126). Histamine-containing mast cells are in close

proximity to the parietal cells, usually in a ratio of 1 mast

cell to every 2-3 parietal cells (119). In the rat, but not

the dog, gastrin increases the mucosal concentration of

histidine decarboxylase, an enzyme that forms histamine

(138). The parietal cell has been shown to be the cell

responsible for histamine uptake and inactivation in the

gastric mucosa (138).

In canine cells, the action of histamine is relatively

weak but is markedly enhanced by the phosphodiesterase


inhibitor, isobutymethylxanthine (138). Species comparisons

have shown that in rabbit cells, concurrent histamine

stimulation is needed before gastrin stimulation will occur,

while in the canine cell and gland, there is a pure gastrin

response coupled to an important histamine component (138).

Histamine is less effective in vivo in stimulating acid

secretion in the rat than in the human, dog and cat (40,107).

Histamine stimulation of acid secretion is mediated by

cAMP and the regulatory subunits of protein phosphokinase

(128,138). Binding of histamine to the H2-receptor on the

parietal cell produces stimulation of adenylate cyclase which

results in cAMP production. This has been confirmed in

studies where H2-receptor antagonists were found to inhibit

production of cAMP during histamine stimulation of canine,

rat and guinea pig parietal cells and rabbit glands. Cyclic

AMP may stimulate a protein kinase which could result in

phosphorylation of protein components involved in HCl

transport, located either in membranes or in the cytoplasm of

parietal cells, such as carbonic anhydrase and the H+-K+-

ATPase. Adenylate cyclase is composed of regulatory subunits

such as the inhibitory guanine nucleotide-binding protein

(138). Prostaglandins have been shown to inhibit production

of cAMP by histamine in canine cells by interaction with the

inhibitory guanine nucleotide-binding protein (128,138). At

one time, histamine was considered the final common mediator

of acid secretion by the parietal cell (33). This theory was


renewed with the development of H2-receptor antagonists which

inhibited gastric acid secretion that had not only been

stimulated histamine, but also by gastrin and acetylcholine

(39,68). There have been 2 types of experiments which tend to

disprove this theory. First, gastric acid secretion was

studied in neonatal rats and occurred only after 2 weeks in

response to only acetylcholine and gastrin (1). An additional

2-3 weeks are needed before a histamine response was

observed. Second, identification of all 3 receptors on the

canine parietal cell also disposes of this once widely held

belief that histamine is the sole final common chemical

mediator for all stimulants (115,138).

Studies in dogs with denervated gastric pouches revealed

that combinations of stimulants produced potentiation,

defined as the simultaneous response to 2 stimulants that is

greater than the sum of the individual responses (40,68).

Potentiation implies that there is convergence on a common

stimulatory pathway after separate receptor binding. In

isolated canine parietal cells, only pairs of stimulants

using histamine or all 3 secretogues together have produced

potentiated responses (138). Potentiation using histamine may

occur because of the central role of cAMP production which

leads to stimulus-secretion coupling (40,138). Cyclic AMP is

responsible for protein phosphorylation which is necessary

for tubulovesicular membrane processes and is required for

ATP production for several ion pumps, namely the Na+- and H+-


K+- ATPases. Neither acetylcholine nor gastrin contributes to

the formation of cAMP nor influences its production by

histamine (138). This absolute component for HC1 secretion

supplied only by histamine may help explain the apparent non-

specific action of H2-receptor antagonists in inhibiting HCl


Non-specific action of H2-receptor antagonists or the

anticholinergic agent, atropine, results in inhibition of

stimulation by any of the 3 secretogues in vivo (40). This

finding suggests that during basal or fasting conditions, the

parietal cells are exposed to a constant background of

histamine and acetylcholine (and possibly gastrin) which

sensitizes the cells to additional amounts of these agents.

If histamine and acetylcholine are present as a potentiating

background, addition of H2-receptor antagonists or atropine

would (and does) interfere with all physiological modes of

acid stimulation (40). This interference does not indicate

whether the antagonist has acted against the basal background

concentration or against additional amounts which might have

been released by the secretogue. For example, H2-receptor

antagonists markedly inhibit acid secretion stimulated by

feeding (33). This suggests that histamine is involved in the

response to feeding but does not indicate whether the

component blocked is only the amount present in the basal

condition or whether feeding has released additional

histamine. In the case of gastrin, which can be measured by


radioimmunoassay, it can be determined whether a certain

mechanism involves additional release of gastrin (104,105).

Currently, methods are not available for determining how much

histamine or acetylcholine within the mucosa is being

released or presented to the parietal cell at a given moment

(40,118). Thus, it is not known how much more of these stores

have been released and are available to stimulate the

parietal cell after any given mode of stimulation.

Phases of Gastric Acid Secretion

Acid secretion has been divided into basal, cephalic,

gastric and intestinal phases although this probably

represents an oversimplification of several interacting

factors (39,68).

Basal Secretion

Basal acid secretion indicates an absence of all

intestinal stimulation and is usually studied after a fast of

several hours to ensure that the intestinal phase of gastric

secretion is over (39). Gastric emptying of a meal is usually

complete by 6 h in a person and by 24 h in a dog (40).

The stimulus for basal secretion is unknown, since it is

decreased but not abolished by vagotomy and antrectomy

(44,66). Histamine may play a role as indicated by the

finding in the rat that depletion of mast cells was followed

by decreased basal secretion but no reduction in response to

vagal and gastrin stimulation (40). Basal secretion in people

displays a circadian rhythm that is low in the morning and


high in the evening (40). This circadian rhythm is not

related to changes in plasma gastrin levels (110).

Concurrent gastric and duodenal motor activity during

basal acid and pepsin secretion has been studied in people

and dogs (64). Vantrappen found in people that phase III

activity of the migrating motor complex in the duodenum was

preceded by an increase in gastric acid and pepsin outputs

(148). In dogs with gastric fistulas, the rate of basal acid

secretion was 2 times higher during phase III activity than

at other times (still only 0.2% of maximal acid secretion)

(139). Thus, concurrent motor activity and gastric secretion

are integrated for digestion.

Cephalic Phase

The cephalic phase of gastric acid secretion is initiated

by stimulation of sensory vagal afferents through the

thought, smell or taste of food (25,141). Uvnas introduced

the theory that vagal impulses could stimulate acid secretion

by 2 mechanisms- acetylcholine released from postganglionic

parasympathetic neurons directly on the parietal cell and

another by stimulation of the G-cell to release gastrin (40).

This theory was confirmed when studies in dogs showed release

of gastrin following sham feeding (66). The maximal acid

secretion rate following sham feeding is 40-100% of the

maximal response to gastrin stimulation (40).

Gastric Phase

The entrance of food into the stomach initiates the second

phase of gastric secretion (39). The gastric phase of acid

secretion was recognized in 1879 by Heidenhain, but the

explanation for this humorally mediated secretion has been

found only recently (40). The gastric phase is initiated by

distension and activation of mucosal sensory afferents.

Distension-mediated acid secretion is thought to involve

stretch receptors which activate both long (vagovagal) and

short (intramural) reflexes involving direct stimulation of

the parietal cells by acetylcholine released from

postganglionic parasympathetic neurons (40,68).

Contact of food with the gastric mucosa also stimulates

acid secretion (39). Chemical agents such as caffeine,

alcohol, calcium ions (human but not dog) and the digestion

products of proteins stimulate chemoreceptors to release

acetylcholine which, in turn, stimulates the release of

gastrin from G-cells in the antrum (39). The digestion of

protein by pepsin preferentially releases tryptophan and

phenylalanine and these 2 amino acids are potent stimulants

of gastrin release (40). The strong correlation between the

increment of serum gastrin concentration and acid secretion

within subjects ingesting peptone meals suggests that gastrin

may be the main mechanism by which amino acids stimulate

gastric acid secretion during the gastric phase (40).

Stimulation of chemoreceptors by protein digestion products


and stretch receptors by distension leads to acid secretion

through acetylcholine-mediated gastrin release and direct

stimulation of the parietal cells by activation of long and

short reflexes.

Intestinal phase

Finally, the entrance of chyme into the small intestine

stimulates the third phase of gastric secretion (39). The

known mediators of the intestinal phase are distension and

protein digestion (39). A humoral mechanism has been proposed

as the response still occurs when all extrinsic nerves have

been cut. Intravenous amino acids stimulate acid secretion so

that part of the mechanism may due to amino acids absorbed by

the small intestine independent of gastrin release (human,


Stimulation of Acid Secretion

The greatest ability of the stomach to secrete acid is

measured by collecting gastric fluid in successive 15-min

periods after the subject has been given a sufficiently

strong stimulus (39). Acid secretion data is generally

expressed in millimoles per unit time and is often expressed

as a fraction of the maximal secretary rate of the stomach

(60,96,125). The maximal acid output (MAO) is the hourly

summation of acid output following a maximally effective dose

of histamine or pentagastrin and is linearly related to

parietal cell mass in people, dogs and pigs (31,100). Maximal

acid output in people has been found to be influenced by sex


(men > women), body weight (larger people have more parietal

cells) and lean body mass (leaner secrete more acid) (39).

A variety of substances have been used to produce maximal

stimulation of gastric acid secretion. Two-deoxy glucose acts

as a competitive antagonist of glucose receptors in the

brain, which stimulates vagal-induced acid secretion in

several species (25,38). In early studies using 2 horses

prepared with gastric cannulas, 2-deoxy glucose (100 mg/kg

IV) inhibited acid secretion while increasing volume

secretion rate and was associated with the side effects of

sweating and bradycardia.

Histamine was used initially to induce maximal acid

secretion in people and animals but required prior treatment

with an antihistamine (Hl-receptor antagonist) to inhibit

systemic side-effects (60,103). The minimal effective dose of

histamine that produced acid secretion in people was 10 times

less than that required for the dog (106). The rat is

insensitive to histamine when compared to the dog, cat, pig,

monkey, or rabbit (40).

Gastrin was used as a stimulant before pentagastrin was

available (60). Pentagastrin (Peptavlon@, Ayerst Labs)

contains the biologically active carboxyl terminal

tetrapeptide of the gastrin molecule and is currently the

preferred agent to study maximal secretary rates (103). The

usual dose of 6 Ag/kg can be given parenterally (usually

subcutaneously). When acid output is expressed in Amoles/kg


body weight to normalize for large differences in size,

pentagastrin and histamine caused similar responses in the

human, while histamine was more effective in the dog and

rabbit (40).

Cholinomimetics are relatively weak stimulants of acid

secretion when compared to histamine and pentagastrin (40).

One possible reason could be that cholinergic agents may

stimulate both inhibitory and stimulatory pathways. Recently,

Feldman and Schiller suggested that the weak stimulatory

effect may be explained by a concomitant stimulation of

nonparietal (bicarbonate) secretion by the surface epithelium


Inhibition of Acid Secretion

Inhibition of acid secretion may be caused either by

direct action on the parietal cell or indirectly by the

release of other inhibitory agents or by changes in blood

flow (119). Gastric acid secretion can be decreased by

inhibition of the H+K+-ATPase, receptor blockage, or

inhibition of intracellular messengers (115). Atropine and

H2-receptor antagonists are potent inhibitors of both

pentagastrin and histamine stimulated acid secretion in the

human, dog and rat (40).

As mentioned earlier, gastrin release can be inhibited by

acidification of the antrum, by release of somatostatin and

by vagal fibers (103). There is evidence that somatostatin is

important in the inhibition of gastrin release by H+ in the


lumen (40,103). Somatostatin cells possess cytoplasmic

processes that extend to neighboring gastrin cells and

parietal cells in the funds. Both the inhibition of gastrin

release and parietal cell secretion are by paracrine and

endocrine effects.

Acid secretion is also inhibited in the intestinal phase

by humoral agents released from the small intestine in

response to duodenal acidification or hyperosmolarity or

increasing lipid content (39,40). A variety of hormones

(gastric inhibitory peptide, enteroglucagon, vasoactive

intestinal peptide and cholecystokinin) are capable of

inhibiting parietal cell response, but their physiological

significance is not established (40). Duodenal acidification

causes the release of secretin from the duodenal mucosa which

stimulates pancreatic bicarbonate secretion and inhibits

gastric acid secretion. Lipids cause the release of

cholecystokinin from the mucosa which stimulates gall bladder

contraction. Cholecystokinin is structurally related to

gastrin in its carboxyl terminus and is a weak agonist of

acid secretion in the cat (40).

Histamine has long been recognized as one of the most

potent stimulants of acid secretion, however, its gastric

secretogue effect is not blocked by conventional

antihistaminic drugs, which inhibit the effects of histamine

on smooth muscle of the intestinal tract and bronchi (33).

This led to the postulation that there were two separate


histamine receptors. Black and co-workers discovered that a

histamine analog inhibited histamine stimulated acid

secretion, identified the receptor on which it acted and

named it the histamine H2 receptor (21).

Since then, numerous H2 antagonists have undergone

clinical trial as potent inhibitors of acid secretion (142).

The first antagonists were developed by lengthening the side

chain and adding a thiourea group to the imidazole ring of

histamine. Cimetidine (Tagamet@, Smith, Kline & French)

contains an imidazole ring and, except for the side chain,

has a structure similar to histamine. Raniditine (Zantac@,

Glaxo, Inc.) contains a furan ring and has been found to be

5-10 times more active on a molar basis than cimetidine in

inhibiting basal and stimulated gastric acid secretion in

humans and dogs (121). Side effects attributed to the

imidazole ring of cimetidine have not been observed for

ranitidine, namely: 1) antiandrogenic effect characterized by

impotence and gynecomastia in men; 2) mental confusion; 3)

interaction with other drugs attributed to inhibition of

hepatic cytochrome P-450 oxidase enzymes; and 4) bone marrow

depression (20,102).

Like cimetidine, ranitidine has been shown to promote

significant healing in human peptic ulceration in several

double-blind clinical studies (142). It is well absorbed from

the human small intestine and has a plasma half-life similar

to that of cimetidine (approximately 90 minutes)


(37,78,92,121). It is excreted in the urine either unchanged

(70%) or as N-oxide and S-oxide metabolites. Ranitidine given

at breakfast to people suppresses acid secretion at lunch and

dinner, thus having a long duration of action. Because of the

increased potency and long duration of effect, ranitidine is

prescribed twice daily and studies are underway to determine

the effectiveness of once per day treatment at night when

acid secretion is highest (142). A possible cytoprotective

effect of ranitidine has been described in people and rats

which may be related to its ability to induce production of

endogenous prostaglandins (151).

Pepsinogen Secretion Rate

The term, pepsin, was introduced in 1836 by Schwann for

the protease in gastric juice (39,74). In 1930, almost one

hundred years later, Northrup crystallized pepsin from

porcine gastric mucosa. Pepsinogen initiates the process of

protein digestion which leads to the formation of peptides

that serve as signals for the release of various hormones,

including gastrin and cholecystokinin (72). In turn, these

hormones serve as major regulators of protein digestion.

Pepsinogen is most active on collagen, which is found in

greater quantity in meat than vegetable protein, implying

that carnivores should have higher secretion rates (72).

Pepsinogen secretion is also thought to be important in those

species where mastication of food is minimal (dog) (72).


Pepsinogens are secreted as a heterogenous group of

inactive proenzymes and are converted into active,

proteolytic enzymes by acid (62,74). Seven distinct mammalian

pepsinogens have been characterized on the basis of

electrophoretic mobility, pH optima and immunologic

reactivity. The five fastest migrating pepsinogens (PI) can

be distinguished from pepsinogens 6 and 7 (P2). Group 1

pepsinogens are found in the chief cells and mucous neck

cells of the fundic glands, whereas group 2 pepsinogens are

found in fundic, pyloric and duodenal glands.

There is a small continuous basal secretion of pepsinogen

in humans (72). Additional pepsinogen secretion is stimulated

by primarily muscarinic cholinergics, but is also stimulated

by pentagastrin and histamine in the human and several other

species (74). In the human or dog, maximal pepsinogen

secretion is 3-4 fold greater than basal rates (72). The rate

of pepsinogen secretion in vivo during histamine

administration in dogs and cats was the subject of

controversy until dose-response curves showed reduced

secretion rates at high doses of histamine (75). Initial

increases, then decreases, in pepsinogen secretion may be due

to release of stored pro-enzyme (wash-out theory) (74).

Pepsinogen secretion activated by luminal acid has also been

demonstrated and probably occurs via local cholinergic

reflexes (72). Differences in species and the combined


possibilities of wash-out and acid activation make evaluation

of pepsinogen secretary data difficult.

After secretion into the gastric lumen, pepsinogen is

irreversibly converted by acid to the active enzyme, pepsin

(Mol. wt. 34,500), by loss of a variable amino terminus

(pepsin inhibitor fragment) (74). Following this loss, the

protein undergoes a conformational change, apparently to

expose a binding cleft that will accommodate a peptide of

about 8 amino acids. The active site contains 2 aspartyl

residues, which is similar to 2 other carboxyl proteases,

cathepsin and rennin. The catalytic site appears similar in

all species, including the horse (62). Once activated, pepsin

can be denatured by pH > 7.2 or temperatures above 65C.

Pepsin is an endopeptidase with a pH optimum of 1-3

depending on the substrate, and is most active against

peptide bonds adjacent to aromatic peptides phenylalaninee

and tyrosine) or dicarboxylic L-amino acids (39,74). Some

free amino acids and polypeptides are produced. The extent of

digestion is determined by the physical state of the ingested

protein, the length of time it stays in the stomach and the

activity of the pepsin. Gastric protein digestion is not

essential in humans, since those with achlorhydria have no

impairment of protein digestion and absorption (39).

The most traditional technique used to determine the

concentration of pepsin in gastric fluid is by the digestion

of hemoglobin substrate at pH 1.7 as first described by Anson


and Mirsky (10). Pepsin splits off products from hemoglobin

which are soluble in trichloroacetic acid. With this

technique, 1 peptic unit (PU) represents the activity of

pepsin which releases 0.1 gM tyrosine from 5 ml of 2%

hemoglobin at pH 1.7 in 10 minutes at 37C (10).

Gastric Fluid Composition

The composition of gastric fluid is determined by the

different secretary systems existing side-by-side in the

gastric mucosa: 1) the parietal cell secretes electrolytes

and acid (HCl); 2) the chief cell synthesizes pepsinogen

which is stored for rapid release but may also be made and

secreted continuously; and 3) the mucous epithelial cell

which constantly secretes water, mucus and bicarbonate (39).

Each secretary product has a specific role in protein

digestion (acid and pepsin) or mucosal protection (mucus and


While pure parietal or nonparietal fluid has not been

measured directly, several theories have been put forth to

explain the composition of gastric fluid in relation to

different mixtures of parietal and nonparietal components.

The ionic composition has been determined for human and

animal species on the basis of positive (or negative)

correlation with the rate of H+ secretion according to the

two-component theory (82,99). This theory assumes that the

parietal component has a constant composition, independent of

the rate of secretion, which is pure, isotonic HCl, and the


nonparietal component is also assumed to have a constant

composition which is an isotonic ultrafiltrate of plasma

containing primarily Na+, bicarbonate, mucus and enzymes

(99). Each component may be secreted at different rates. At

maximal rate of acid secretion, gastric fluid is an isotonic

mixture of almost pure HCl mixed with K+ (estimated ionic

composition: 149 mM H+, 166 mM Cl-, 17 mM K+) (84,98). The

composition of the nonparietal component has been estimated

to be 140 mM Na+, 5 mM K+, 120 mM Cl-, 25 mM HCO3. To

maintain the striking gradients between H+ and Na+ between

plasma and gastric fluid, the gastric mucosa is practically

impermeable to the passive diffusion of these ions and

maintains a transepithelial potential gradient of 40 mV

mucosaa negative) (39). Sodium is apparently not essential

for Cl- entry into the cell and has a constant output with an

inverse relationship to H+ concentration (84,98).

In addition to active H+ secretion, chloride has been

shown to be actively secreted by the gastric mucosa against

both a concentration and electrical gradient (95,119). This

active secretion results in a negative potential difference

mucosaa negative). The existence of the potential is not

dependent on the acid secretary rate of the mucosa, as the

stomach secretes Cl- even when there is no H+ secretion, but

is produced by active Cl- secretion in excess of the H+

secretion (95,119). This so-called non-acidic Cl- secretion

manifests itself as the lumen-directed negative short circuit


current and PD. Measurement of Cl- fluxes in amphibian or

piglet mucosa clearly shows that upon stimulation of acid

secretion the Cl- transport from blood to mucosa increases

with a value corresponding to the H+ secretion. Passive and

facilitated Cl- transport systems have been identified and

consist of the following systems: 1) an anion-coupled

antiport Cl-/HCO3- exchange at the basolateral surface; 2)

Na+ facilitated Cl- transport at the basolateral surface

(uncertainty if NaCl co-transport or Na+/H+ and Cl-/HCO3-

exchange); 3) a KCl co-transport at the apical surface; and

4) Cl- conductance pathways at the apical and basolateral

membranes (95).

Gastric secretion of bicarbonate ion was postulated by

Schierbeck in 1892, and Hollander proposed that secretion of

a bicarbonate-containing fluid was the main mechanism for

regulating gastric acidity (52,82). Gastric bicarbonate has

only recently been found to be an active process which can be

stimulated and inhibited in the fundic, pyloric and duodenal

mucosa (52,46). The amount of bicarbonate transported by the

mucosa might be sufficient to account for the well known

constant loss of H+ from acid contents in accordance with the

two-component theory. The amount of Na+ co-transport or

reaction with H+ would decrease luminal osmolality and excess

water would move into the mucosa (51). However, bicarbonate

secretion may also appear as a surface neutralization of H+

rather than as a secretion with a defined volume (53). Direct


measurements of pH at the mucosal surface strongly suggest

that neutralization of H+ diffusing from the lumen by

bicarbonate from the epithelial surface occurs in a narrow

zone within the surface mucous gel layer and not in the

luminal bulk solution (52). The exact mechanism by which

gastric bicarbonate is secreted is unknown. The fact that

there is virtually no change in gastric electrical potential

difference during stimulation of bicarbonate secretion

suggests that bicarbonate transport may take place via an

electroneutral ion exchange mechanism (such as with Cl-) at

the luminal or basolateral membrane (95).


Peptic Ulceration

Gastric Ulceration

Although gastric ulcers in foals have been reported since

1964, information is still lacking on the prevalence of

squamous versus glandular lesions (15,30,124). Erosions and

ulcers are found usually at the margo plicatus in the

squamous mucosa and have been reported as incidental findings

in 2-25% of foals presented for necropsy. Fifty percent of

perforated ulcers in one study were in the squamous mucosa.

These lesions have also been recently reported in adult

horses without a history of previous medication. Diet and

physical stress have been proposed as etiological factors.

Ulcers in the squamous mucosa near the cardia (junction of

esophageal and gastric mucosa) are common in swine fed high

protein or pelleted diets. Surveys of farms with equine

gastroduodenal ulceration have not shown any effect of diet

but additional studies are needed concerning feeding

practices on these farms.

The squamous mucosal ulcers in foals may be similar to

"stress erosions" seen in seriously ill people following

major surgery or trauma (103). Stress erosions in people,


however, occur in the glandular mucosa. The pathogenesis of

gastric ulceration in people, in general, is less associated

with excess acid and pepsin secretion and more associated

with decreased mucosal resistance normally produced by

secretion of mucus and bicarbonate. With restoration of

normal defense mechanisms, stress ulcers heal rapidly within

several days. Experimental evidence in laboratory animals

also suggests that stress ulceration results from a temporary

failure of gastric mucosal defense mechanisms, rather than

from excessive secretion of acid and pepsin, however, mucosal

lesions can not be produced without luminal acid (103). Since

the equine squamous epithelium does not appear on histologic

sections to be covered by mucus to as great an extent as the

glandular epithelium, it may be more easily damaged than the

glandular epithelium when there is failure of gastric mucosal

defense mechanisms.

Equine gastric ulcers of the glandular epithelium are seen

less frequently than those of the squamous mucosa and have

been reported in ill neonates (30). Glandular gastric ulcers

have also been experimentally produced with nonsteroidal

anti-inflammatory drugs (NSAID),in foals and ponies (145).

Ulceration following NSAID administration probably results

partly from inhibition of endogenous prostaglandin synthesis.

Prostaglandins have been shown to inhibit acid and pepsin

secretion, stimulate mucus and bicarbonate secretion and

maintain mucosal blood flow (122,123,151). Additionally,


prostaglandins promote rapid restitution of damaged

epithelium in doses which do not inhibit acid secretion


Duodenal Ulceration

The equine cranial duodenum is sigmoid-shaped with an

initial ampulla separated by a slight constriction from the

second ampulla into which the hepaticopancreatic duct enters

(135). Duodenal ulcers in the horse are generally located on

the anti-mesenteric surface as solitary or multiple lesions

in the proximal duodenum or can diffusely involve the entire

duodenum (15,28). The hepatic duct may be secondarily

obstructed due to duodenitis or stricture formation distal to

its opening into the duodenum.

Duodenal ulcers are more commonly seen in 2-6 month old

foals and can occur as a herd outbreak (30). Diarrhea

frequently precedes other clinical signs which include teeth

grinding, excessive salivation, depression, anorexia and

colic. Delayed gastric emptying usually results in squamous

gastritis and esophagitis. In foals with gastric outlet

obstruction from duodenal ulcer disease, gastrojejunostomy

has resulted in rapid healing of gastritis and esophagitis

without recurrence of duodenal ulceration (28).

Unfortunately, duodenal ulcers in foals may perforate and

result in rapid life-threatening deterioration within 48

hours of the first clinical signs of ulcer disease.


Human duodenal ulceration represents the classic condition

brought about by the proteolytic action of acid and pepsin

(103). Excess acid may result from many factors such as an

increased basal or nocturnal secretion rate, increased

parietal cell number, increased vagal inputs in response to

food, increased sensitivity to or release of gastrin and

heredity. Breakdown of mucosal resistance is also important

in duodenal ulceration and a decreased bicarbonate secretion

rate has been reported (49).


The mechanisms by which the gastric and duodenal mucosa

normally resists the onslaught of food and secreted acid and

pepsin are not completely understood (39). It is generally

thought that mucosal damage is induced by acid, pepsin and

bile after the integrity of the gastroduodenal mucosa has

been damaged. Research into the etiology of human duodenal

ulceration has tended to concentrate on factors that control

acid and pepsin secretion, while factors that determine

mucosal resistance have been investigated in gastric

ulceration (103). It is generally agreed that the presence of

sufficient luminal acid and pepsin are required for both

gastric and duodenal ulceration to develop. Gastric acid is

considered the primary damaging agent which also activates

the proteolytic enzyme pepsin from secreted pepsinogen.

Mucosal ischemia appears to compromise the ability of the

gastric mucosa to dispose of intracellular acid which may


have diffused into the mucosa (39). Although gastric acid and

pepsin seem essential for the formation of ulcers, gastric

ulcer patients frequently secrete acid at rates

indistinguishable from normal patients (103).

In contrast, nearly half of the duodenal ulcer patients

have increased basal and maximal acid secretion rates (103).

The mechanisms for these augmented secretary rates have been

attributed to increased responsiveness of the gastric

parietal cell to stimulation by neural and hormonal (gastrin)

agents, increased parietal cell mass, hypergastrinemia and

impaired duodenal mucosal bicarbonate secretion. The

pathophysiological processes underlying duodenal ulceration

are not fully understood but there are indications that one

of them may be an inadequate neutralization of an excessive

gastric juice entering the duodenal bulb. The rate of acid

and pepsin delivered to the duodenum by factors regulating

gastroduodenal motility are also being investigated with

regard to development of peptic ulceration.

In circumstances of normal secretary rates, acid probably

plays primarily a permissive role and attention has focused

on factors that may protect the mucosa from acid and pepsin.

The components of the mucosal barrier include a surface layer

of mucus, surface epithelial bicarbonate secretion, mucosal

blood flow, epidermal growth factor concentration (inhibits

acid secretion and epithelial repair and regenerative

processes (30,39,114). Agents known to "break" the barrier


are acids (HCl, salicylic), detergents that attack the lipid

portion of the plasma membrane (bile acids), hyperosmolar

solutions and ethanol (39).

Mucus is secreted as an insoluble gel adherent to the

mucosa and as a soluble form mixed in the lumen with gastric

contents (149). It functions as a barrier for hydrogen ions

with neutralization by bicarbonate secreted from the surface

epithelium, as a diffusion barrier for pepsin and as a

lubricant to prevent mechanical damage (14,69,152). Finally,

the surface epithelium is capable of rapid restitution after

acute superficial damage by migration of the surface


Secretion of bicarbonate by both gastric and duodenal

surface epithelium beneath the mucous layer maintains the

surface pH near neutrality in spite of a luminal pH as low as

1 (52). Bicarbonate secretion is stimulated by luminal acid

and this response is mediated by endogenous prostaglandins,

humoral factors and, probably, neural inputs (47).

Abnormalities of gastroduodenal motility have been

investigated in people with peptic ulceration (30,103). In

both gastric and duodenal ulceration, increased

duodenogastric reflux has been reported. Delayed gastric

emptying can produce gastric ulcers by increasing the

duration of mucosal contact with acid and pepsin. Delayed

gastric emptying can result from gastric stasis due to

gastroenteritis or can be secondary to duodenal ulceration.


In people with gastric ulceration, antral hypomotility, and

hence decreased emptying of solid foods, has been reported.

In duodenal ulceration, gastric accommodation to distension

is not altered nor are there changes in the rhythm or

frequency of gastric peristalsis. There are reports of

increased gastric emptying rates and the rapid entrance of

acidified chyme into the duodenum could contribute to the

inability of the duodenum to neutralize the acid, leading to

ulceration. An abnormality in local motility of the duodenum

might contribute to the formation of ulcers at specific

sites, such as the duodenal bulb, as the proximal duodenum

helps to regulate emptying of acid from the stomach. Delayed

gastric emptying has been reported in foals with duodenal

ulceration and it has usually been a result of stricture

formation from healed ulcers. Information is not available

concerning changes in gastric secretion or motility in these

foals, which could provide clues to the pathogenesis of this

disease in horses.

Inhibition of endogenous prostaglandin synthesis by NSAID

seems likely to be the cause of mucosal damage that can be

produced by these agents (123,151). Some NSAID's also

increase gastric acid secretion. Prostaglandins inhibit the

stimulation of cAMP production from ATP by inhibition of

adenylate cyclase (138). This enzyme is stimulated in the

gastric parietal cell following binding of histamine as

mentioned earlier. Several experimental studies in foals and


ponies have shown ulcerogenic properties of phenylbutazone

(30,144). Oral lesions were observed more frequently

following oral administration than parenteral. Administration

of an oral prostaglandin E2 prevented phenylbutazone-induced

ulceration in ponies, although it is not known if the dose

used also decreased acid secretion.

An infectious or toxic etiology has been proposed to

explain the occurrence of diarrhea and clustering of affected

foals on certain farms (15,30). Such an agent could directly

damage the gastroduodenal mucosa or possibly affect

gastroduodenal motility, resulting in ulceration. An organism

or toxin has not been isolated (2). Currently there is much

excitement concerning the possible role of Campylobacter

pyloridis in the pathogenesis of human gastric ulceration

(26). Campylobacter sp. are urease-positive and lie beneath

the antral mucosa. The bacteria might damage the mucosa by

trapping ammonium thereby exposing the mucosa to excessive

hydrogen ions.

Bile salts have been implicated in causing gastric mucosal

damage and have also been found to inhibit mucus and

bicarbonate secretion in people (30). With recent techniques,

bile reflux has been found in normal fasting humans, dogs and

pigs, and has been observed during endoscopy and

experimentation in normal horses. Bile reflux may be

important if the gastric mucosa is already damaged. In foals

treated for gastric outlet obstruction with


gastrojejunostomy, clinical signs of ulceration did not

worsen after total diversion of proximal duodenal contents

including bile into the stomach. Bile reflux may not be a

major factor in the development of gastric ulceration in the


There has been interest in pepsinogen as a contributor to

the formation of peptic ulceration (149). Pepsin constantly

digests the mucous layer covering the mucosa. Several studies

have shown that the instillation of acid alone was

insufficient to produce ulceration; pepsin had to be present

as well. In duodenal ulcer patients, both basal and

stimulated pepsin secretion, like acid secretion, are usually

higher than in normal subjects. Both pepsinogens 1 and 2 can

be detected in serum and Samloff et al. found a greater

proportion of P1 in duodenal ulcer patients than normal

subjects (129). Pepsinogen 1 has a mucolytic activity over a

wider range of pH including that seen in the duodenal bulb.

However, there is considerable overlap in serum pepsinogen

concentrations between normal subjects and patients with

duodenal ulcer disease and thus serum levels are not

considered diagnostic. The lack of interest in antipepsin

therapy is due in part to the success of antacid therapy,

since acid secretary inhibitors suppress pepsinogen secretion

and lead to a higher luminal pH which reduces pepsinogen



Equine Gastric Secretory Physiology

Until recently, equine gastric physiology has received

little interest primarily due to a paucity of gastric

diseases and difficulties in creating a gastric fistula model

in this species. Little information is therefore available on

the composition and control of gastric secretion in the

horse. Russian investigators described experiments in fasted

young horses (1.5-2 years old) with a gastric fistula and

found a continuously variable gastric fluid output that could

not be further stimulated by offered food (73,145). Atropine

caused regurgitation of duodenal contents and markedly

decreased the volume of secretion. The pH of equine gastric

contents has been reported to range from 1.13-6.8 and 1.6-6

(necropsy specimens).

Volatile fatty acids and lactic acid were measured in

gastric contents of the fed horse (8,13). Equine fundic and

pyloric mucosa were found to absorb volatile fatty acids but

only the pyloric mucosa transported significant amounts to

the serosal surface (13). The stratified epithelium did not

absorb the fatty acids and had a low tissue conductance

indicating it was relatively impermeable to passive ion

transport. The nonglandular and glandular regions had similar

pH recordings and the lowest values obtained after feeding

had a mean value greater than pH 2 (13). Electrolytes have

also been measured in the gastric contents of fed ponies, and


had mean values of 62 12 mEq/L Na+, 28 5 mEq/L K+, 78

15 mEq/L Cl- and 8.6 2 for total 002 (5,11).

Gastric empyting is rapid in the horse as judged from

radiographic studies and from studies using liquid and

particulate markers (7,12,27). Gastric emptying rate in foals

was studied by contrast radiology and has been reported to be

slower in the nursing foal than in the weaned foal in one

study (7). However, another study did not find any difference

in gastric emptying rate due to age (27). In the adult, 75%

of a liquid marker had emptied by 30 minutes; however

nondigestible particulate markers were retained for a length

of time directly proportional to their size (12). Gastric

capacity is estimated to be only 10% of the total

gastrointestinal tract (135).

A technique for gastric cannulation has only recently been

reported and was developed by the authors for these studies

(29). Equine gastric fluid was secreted continuously during

basal conditions and acid output was inhibited by ranitidine

given at 0.5 mg/kg IV. Although several other investigators

have adapted the use of this cannula technique, results of

their studies have not yet been published.


Animals and Surgical Preparation

Thirteen horses of various breeds and sexes, ranging from

4-12 months of age, were selected since the problem of

gastroduodenal disease is seen primarily in 2-6 month-old

horses (Table 3.1). The horses were free of clinical signs of

gastrointestinal disease and were dewormed (Eqvalan@, Merck &

Co., Inc.) and vaccinated for encephalitis and tetanus

(Cephalvac@, VEWT, Cooper Animal Health Inc.) before surgery.

The gastric cannula was constructed entirely of medical

grade silicone rubber tubing (1.6-cm, ID; 2.2-cm, OD;

Silastic, Dow Corning Corp, Midland, Mich). Three flanges

were made from the silastic tubing and two were permanently

fixed to the tube with silicone glue, one at the inner

gastric end and the other at the inner ventral body wall

position. The middle flange (outer gastric flange) was freely

movable to allow positioning after placement of the tube with

the inner flange within the gastric lumen. The cannula was

also supplied with an inner tube plugged at the gastric end

to occlude the cannula between experiments. The inner tube

and an outer body wall flange made of soft rubber were held

in place with a hose clamp.

Table 3.1. Animal signalment.

Horse # Age Sex Breed Weight

(m) (kg)

13 6 M

Quarter Horse

Quarter Horse







Quarter Horse


Quarter Horse

Quarter Horse

Quarter Horse




























The horse was anesthetized with halothane and a ventral

midline incision made to gain access to the cranial abdomen.

The stomach was isolated with moistened towels to prevent

contamination of the abdomen with spilled gastric contents.

Two stay sutures were placed about 20 cm apart on the ventral

surface of the stomach between the lesser and greater

curvatures and were used to elevate the stomach to the

incision. Three purse-string sutures were placed around the

intended gastrotomy using #2 polyglactin 910 (Vicryl, Ethicon

Inc.). A stab incision was made in the center of the purse-

string suture pattern and the cannula was inserted into the

gastric lumen. The sutures were tied allowing the serosa to

invert around the tube before securing the outer gastric

flange to the serosa with #2 polyglactin 910. The cannula was

brought through a stab incision in the left ventral abdomen

and the abdominal incision was closed with 3 layers of suture

in a routine manner.

Horses were allowed to drink only water during the first

24 hours after surgery and then were gradually fed alfalfa

hay. Each horse was given gentamicin (2 mg/kg IM TID) and

procaine penicillin (10,000 U/kg IM BID) perioperatively and

a nonsteroidal anti-inflammatory drug (flunixin meglumine,

Banamine@, Schering Corp.) was administered if any signs of

pain were observed.

The horses were given 2 weeks to recover from surgery and

were trained to accept gentle restraint by cross-tying in a


stocks. Body weights were determined biweekly. They were

maintained on a diet of grain and free choice coastal hay and

salt block. After several weeks, they could be turned out in

a large pasture to exercise.

Endoscopy was performed via the cannula to evaluate the

condition of the gastric and duodenal mucosa in horses 5-13

(Video-Endoscope@, Welch-Allyn). Each horse was euthanitized

after completion of the studies and a full necropsy was

performed. Histologic sections were examined from the gastric

squamous and glandular mucosa, margo plicatus and duodenal

mucosa using hematoxylin-eosin and PAS stains.


Experiments were designed to address the following


1) Protocol I: What is the basal acid secretion rate in

the horse?

2) Protocols II and III: Can the equine stomach be

stimulated to secrete acid by intravenous pentagastrin

infusion and what is the dose which will produce maximal acid


3) Protocol IV: What are the effects of a H2-receptor

antagonist (ranitidine) on equine gastric acid and pepsin

secretion under basal and pentagastrin stimulated conditions?

Secretory Experiments

The horse was fasted for 24 hours with access to water and

then placed in a room isolated from other horses and food.


The experiments began between 7:30 and 9:00 am to control for

possible diurnal variation in acid secretion as has been

described for people, but were not addressed in this study

(48). They were conducted once per week to control for

possible residual drug effects and to maintain animal health.

Isotonic saline was administered at 200 ml/h for 5 hours

through a jugular catheter to replace electrolytes lost

during the experiments and the catheter was used as the route

of administration for test drugs. A plasma sample was taken

at catheter placement for determination of plasma osmolality

(all horses). The gastric cannula was opened for 30 minutes

to allow gravity drainage of any residual chyme before

beginning the 5 hour collection.

Experimental Design

Protocol I

Gastric fluid was collected every 15 minutes for 5 hours

in 13 horses to determine basal rates of secretion.

Intravenous 0.9% saline was administered at 200 ml/h.

Protocol II

Gastric fluid was collected as for protocol I. Two horses

were given step-wise increasing doses of pentagastrin

(Peptavlon, Ayerst Labs) at 1.5, 3, 6 and 12 4g/kg/h by

intravenous infusion (Harvard infusion pump, Harvard

Instruments). The pentagastrin was dissolved in water by

increasing the pH to 9 with 0.1 N NH4OH, and NaCl was added

to achieve isotonicity before a final pH adjustment to 7.4. A


one hour baseline collection was made and the pentagastrin

infusion started at the beginning of the second hour. These

doses of pentagastrin were selected based on studies in other


Protocol III

Gastric fluid was collected as for protocol I. At the

beginning of the second hour, 8 horses received pentagastrin

at 6jg/kg/h by intravenous infusion.

Protocol IV

Gastric fluid was collected as for protocol I. At the

beginning of the second hour, 4 treatments were given (Table

3.2). A randomized block design was used where five horses

received 4 treatments in random sequence and in duplicate (8

experiments per horse). The treatments were randomized to

minimize effects due to the gastric cannulation (fatigue) and

adaptation by the animal to the secretary studies (learning).

As each horse served as its own control, individual variation

could be evaluated and since duplicates of each treatment

were performed, any missing data was accounted for by using

an average of the two treatments. The use of repeated

measures also increased the confidence in the results given

the small number of animals.

Measured parameters

The fluid was collected in a Erlenmeyer flask, and every

15 minutes the flask was emptied into a graduated cylinder.

The volume was measured to the nearest 2 ml and the sample

Table 3.2 Treatment scheme for protocol IV.


1 2 3 4 5


1. Salinea- -------- ------- -------

2. Saline----------------- ----------------------- -

3. Saline----------------- -------------------------
Pentagastrinc- -------- ----------

4. Saline----------------- ------------ ------------
Pentagastrin-------------- -----------

a NaCl 0.9%, 200ml/h.
b Ranitidine IV bolus, 0.5 mg/kg.
c Pentagastrin infusion, 6 jg/kg/h.


was decanted to remove a top layer of foam (usually 5-20 ml).

The sample was filtered through glass wool to remove any

particulate material and an aliquot saved.

Each aliquot was analyzed the same day for pH, H+

concentration and osmolality. The pH was measured immediately

using a glass electrode with the pH meter (PHM 61,

Radiometer) calibrated at 20C using commercial buffer

solutions of pH 2 and 7 (pH Standard, Fischer Scientific

Co.). Hydrogen ion concentration (H+ mEq/L) was determined in

duplicate by titrating a 1 ml sample diluted 1:10 with

deionized water with 0.1 M NaOH to pH 7.4. The number of

milliliters of titrant was equal to the molarity*10-1 of HCl

in the gastric juice (ex. 2 ml of 0.1 M NaOH = 1 ml of 0.2 M


Gastric fluid and plasma osmolality were measured in 200

uL samples by freezing point depression (Advanced Digimatic

Osmometer, model 3DII, Advanced Instruments, Inc.) and

expressed as mosmoles per kilogram.

An aliquot from alternate time periods (ie. every 30

minutes) was stored at 4C for pepsin determination which was

performed within 1 week, using a modification of the Anson

and Mirsky method (19). All reagents were equilibrated to

25.5C (method 1) or 37C (method 2). Bovine hemoglobin

substrate (3 ml of a 2% solution in 0.06 M HCl; Fischer

Scientific) was incubated with a 0.5 ml sample of gastric

juice diluted 2.5-10 times with 0.01 M HCl for 10 minutes and


the digestion stopped by addition of 5 ml of 5%

trichloracetic acid. The digestion mixture was agitated and

filtered through #2 or #50 Whatman paper. Peptic activity was

estimated in the resulting filtrate by measuring the density

of color produced by liberated tyrosine and tryptophane

(phenol reagent method, method 1) or by their absorbance at

280nm (UV method, method 2). The phenol method determines the

content of split products in the resulting filtrate by

addition of 0.5 M NaOH and phenol reagent and reading

absorbance at 578 nm. Pepsin concentration (Ag/ml) was

determined from a standard curve based on 5-40 Ag of

crystalline pepsinogen (Cooper Biomedical). A blank was

prepared for each sample by precipitation of the hemoglobin

substrate with trichloroacetic acid before addition of the

sample. The 2 methods were compared on 3 occasions by

preparation of standard curves from the same pepsin

standards. Coefficients of correlation between the 2 methods

were determined as well as the mean, standard deviation and

coefficient of variation for replicates of standards.

Calculated parameters

Acid (H+ gEq/kg) or pepsin (pepsin gg/kg) output was the

product of the respective concentration and volume. Output

values were divided by the animal's body weight to normalize

the data between animals and to allow comparison with other

species of various sizes. Basal acid output (BAO, gg/kg/h)

was the sum of the first four 15-minute periods during which


no treatment other than intravenous saline had been given and

was calculated for each experiment. Maximal acid output (MAO,

jg/kg/h) was the sum of the four consecutive 15-minute

periods with the highest total after pentagastrin infusion.

Maximal acid concentration (MAC) was the highest hydrogen ion

concentration achieved during pentagastrin infusion.

Statistical Analysis

Means and standard deviations (STD) were calculated for

all variables (volume, pH, hydrogen ion concentration, acid

output, pepsin concentration and output, osmolality) with n

equal to the number of animals. Correlation coefficients were

calculated for overall and hour 1 control gastric fluid

variables and between the 2 methods for pepsin analysis.

Coefficients of variation (CV, STD/Mean*100) were calculated

within- (n=10, average of replicate experiments coefficients

of variations) and between-animals (n=13) during control

experiments. Within-assay coefficients of variation were

calculated for replicates of hydrogen ion concentration,

osmolality and pepsin concentration.

The hourly average for each variable was analyzed by

repeated measures ANOVA with 2 within-subject factors,

treatment (4 levels) and time (5 levels). If the treatment-

time interaction term was significant, a multiple comparison

test was used to determine significant differences between

means. Both within-treatment (ie. treatment 1, hours 1-5) and

within-time (ie. hour 1, treatments 1-4) were tested. The


Scheffe test is considered a conservative multiple comparison

procedure and was used for all variables except for pH, where

a t-test with pooled variances was used. The t-test is a more

liberal multiple comparison test and was used because of the

large standard deviation in control pH and small differences

between treatment means. Significance was determined at P <

0.05 for both tests.

Phenol Red Recovery

The efficiency of gastric fluid collection was tested by

instilling a phenol red solution (phenolsulfonphthalein,

Sigma Chemical Co.; 50 mg/L in 0.16M HCl) over a 1-minute

period into the stomach of each fasted horse (Trial 1). Five

to ten weeks later, the test was repeated (Trial 2). Gastric

contents were collected at 1, 5, 15 and 30 minutes and

volumes were recorded. Colorimetric estimation of phenol red

concentration was determined in each gastric fluid sample

after a 1:5 dilution and alkalinization with 1M NaOH.

Absorbencies were measured spectrophotometrically at 520, 560

and 600nm and a 3-wave length correction applied, correcting

for background absorbencies in this region of the spectrum


[Phenol Red] = 560 0.5 (520+600)

Phenol red recovery rate (%) was determined by summation of

each sample's phenol red content divided by the total amount



Animal Preparation and Necropsy Findings

All horses recovered from surgery without major

complications. In several animals, fever developed 3-8 days

post-operatively which was attributed to focal inflammation.

Abdominal pain occurred in most animals, but was infrequent

and readily managed with xylazine (Rompun@; Haver-Lockhart)

or flunixin meglumine.

One horse was euthanitized after experiencing several mild

bouts of colic 10 weeks after surgery, and necropsy findings

revealed massive abdominal adhesions from generalized

peritonitis. Another horse developed a fistula between the

left dorsal colon and stomach along the cannula. In the other

horses, the only remarkable findings on necropsy were fibrous

adhesions around the cannula with adhesion of the large colon

to the ventral body wall. Few adhesions were found in the

cranial abdomen indicating peritonitis was not a major

complication of this procedure. In animals cannulated for

over 2 months, the cannula became incorporated into the wall

(2) or lumen (5) of the large colon at the diaphragmatic or

sternal flexure. This incorporation occurred without apparent

abdominal pain or abscessation.


The gastric mucosa was in good condition in all horses on

gross inspection. In one horse, a glandular gastric ulcer,

discovered by endoscopy 12 weeks prior to necropsy, was

completely healed. It was located opposite the cannula on the

lesser curvature and may have resulted from using a long

inner-tube for the cannula. Six horses had 1-3 small squamous

mucosal erosions at the margo plicatus. These lesions

appeared histologically as a mild superficial gastritis with

primarily a neutrophilic infiltrate. Two horses had 1-2

linear mucosal creases at the gastrotomy which were

associated with the inner-most suture line. The duodenum

appeared normal in all animals.

Phenol Red Recovery Rates

The results of phenol red recovery are shown in Table 4.1

and demonstrated that most horses had greater than 80% of the

instilled solution recovered through the gastric cannula. The

majority of the solution was recovered by 1.5 minutes. A

greater recovery rate was achieved by administration of the

larger dose in horses 3-13. There was generally good

agreement between duplicate trials in horses, and also

between recovery rates at the beginning and end of an

experiment. The corrected standard curve was linear.

Gastric Secretory Studies

Results from the gastric secretary studies are described

for each variable during control (saline) and stimulated

(pentagastrin) treatments. For example, all control

experiments whether performed for protocol I or IV, will be

presented together. This has been done to allow the most

accurate estimate of each variable with the greatest sample

size. Finally, for protocol IV, the results are analyzed by

variable for each treatment (control, stimulated, inhibited,

stimulated then inhibited) and time (hours 1-5) (Analysis of

variance, 2 within-subject factors). Data are reported as the

mean standard deviation (STD, n= number of horses), except

in the figures where the standard error of the mean (SEM) is

shown. A P value of less than 0.05 was considered

significant. This statistical analysis allowed each variable

Table 4.1. Phenol red recovery rates (%).

Horse Trial 1 Trial 2 AVG.

1 45 51 48
2 89 64/79b 84
3 100 92 96
4 92 ND 92
5 95 97 96
6 91 81 86
7 87 ND 87
8 86 100/83b 92
9 76 ND 76
10 98/106b ND 102
11 82 91 87
12 82 ND 82
13 75 ND 75

a Solution administered at 500 ml
except horses 1 and 2 (200 ml).
b Before and after an experiment.


to be analyzed with regard to the great amount of individual

horse variation in basal secretary rates.

Control Studies

Collection of gastric contents for 5 hours did not result

in dehydration or substantial electrolyte loss for any of the

horses as packed cell volume and total protein and plasma

sodium, potassium and osmolality did not change. After a fast

of 24 hours, the gastric contents contained chyme (100-500

ml) which was removed during the initial 30 minute drainage

period before the 5-hour study began. Gastric fluid was clear

and watery in samples with a low pH and was dark green to

yellow and viscous in samples with a high pH. Periodic

duodenal reflux, indicated by a high pH and visible bile

discoloration, occurred intermittently in most horses. After

the fluid was measured in a graduated cylinder, a white

frothy layer floated to the top (5-30 ml) in most collections

and was presumed to be saliva as the pH was greater than 8.

Basal gastric fluid secretion was determined in 29

experiments in 13 horses (Table 4.4, Figure 4.2, Appendix A).

Gastric volume (flow rate), pH, H+ concentration (H+ mEq/L)

and acid output (AO, H+ gEq/kg) were continuously variable

between 15-minute collections. Gastric flow rate ranged from

50 to 710 ml/15-min with a mean of 290 80 ml/15-min (STD).

Gastric pH was highly variable and ranged from 1.3 to 6.4

with a mean of 1.98 0.51.


Hydrogen ion concentration and acid output varied markedly

between and within horses. Hydrogen ion concentration ranged

from 0 to 67 mEq/L with a mean of 39 15 mEq/L and acid

output ranged from 0 to 131 /Eq/kg/15-min with a mean of 56

30 AEq/kg/15-min. Basal acid output (BAO) was determined by

summation of the first 4 15-minute periods and ranged from 0

to 541 AEq/kg/h with a mean of 210 109 AEq/kg/h.

The results are expressed both as an overall average of 20

15-minute collections and as the average of the first 4 15-

minute collections (hour 1) in Table 4.2. Average values of

the first hour were highly correlated to the overall average

(Appendix C).

Pentagastrin Stimulation Studies

Protocol II

Pentagastrin, given by step-dose infusion (1.5, 3, 6 and

12 Ag/kg/h), produced an augmentation in acid output by

increasing both hydrogen ion concentration and secretary rate

in 2 horses (5 experiments; Appendix A). Six micrograms per

kilogram was selected as the dose producing maximal

stimulation of acid output, as doubling the dose decreased

acid output (Table 4.3).

Protocol III

Pentagastrin, 6 Ag/kg/h by intravenous infusion, produced

increases in hydrogen ion concentration and volume output

which reached a steady and maximal rate by the second hour in

14 experiments in 8 horses (Figure 4.2, Appendix A). Acid

Table 4.2 Control gastric fluid variable averages
(29 experiments in 13 horses).

Overall (5 Hours)
Coefficient Variationa
Mean STD Inter Intra

Volume (ml/15-min) 287 77 27 19

pH 1.98 0.51 26 24

H+ mEq/L 39 15 39 27

H+ AEq/kg/15-minb 56 30 54 33

Hour 1

Volume (ml/15-min) 295 84 28 20

pH 2.03 0.58 28 26

H+ mEq/L 41 26 32 47

H+ AEq/kg/15-minb 52 27 53 39

a Inter-animal coefficients of variation for 13 animals
and intra-animal coefficients of variation based on
average coefficients of variation for replicate
experiments in 10 animals (see Appendix A for number of
experiments per horse).
b Acid output per period.





I I-_
I *-" --- i \ U
{ I I Em. L !" "-
,\,-"- ~ ~~~~ I 1 |*...



Mean SEM

Time (Hour)

Figure 4.1. Basal acid output in 13 horses (29 experiments).
All horses received 0.9% NaCI, 200 ml/h IV, throughout the 5-
h experiment.

! !


Table 4.3 Maximal acid output during step-dose
pentagastrin infusion for 5 experiments in 2
horses (mean STD).

Dose (jg/kg/h) Maximal Acid Output (H+ AEq/kg)
per 15-minutes per hour

1.5 128 27 338 53

3.0 188 28 496 94

6.0 197 31 495 104

12.0 145 19 455 60



100 1 T T T T T

80 IiI

S 60 Tj
+ -
T 40-i

20- Mean SEM

0- i i i i-
0 1 2 3 4 5
Time (Hour)

Figure 4.2. Acid output following pentagastrin infusion (6
gg/kg/h), which began at hour 2 (14 experiments in 8


output was increased within 30 minutes. The maximal hydrogen

ion concentration (MAC) achieved was 97 mEq/L (pH 1.15),

while the average MAC was 65 18 mEq/L. Gastric fluid pH was

lowered to 1.3 0.15 and had less variation than during the

control experiments (coefficient of variation = 7%).

The maximal acid output (MAO) during the second hour of

infusion ranged from 296 to 944 gEq/kg/h, with an overall

mean of 622 171 gEq/kg/h. Analysis of variance indicated

that during the second and third hours of pentagastrin

infusion, acid output was maximal and greater than the first

and fourth hours (Appendix B). The coefficients of variation

for MAO were approximately 25% for each hour, which were much

less than that observed during the control hour (57%).

Pentagastrin administration produced excessive salivation

in one horse. Several horses started grinding their teeth,

but this was also observed during experiments without

pentagastrin. One horse developed mild gastric hemorrhage

during the last 30 minutes of pentagastrin infusion, and

diffuse glandular mucosal hyperemia and petechiation were

seen on endoscopy.

H2-Receptor Antagonist Studies

Protocol IV

Each response variable was analyzed by repeated measures

ANOVA within each treatment, followed by a within hour

analysis (Appendix B). This allowed each variable to be

evaluated both to its own daily control (HR1) and to the


other treatments at the same hour. Due to the small sample

size, only those P-values greater than 0.05 by Scheffe's

multiple comparison procedure are reported, as well as

results following the Tukey HSD and t-test in Appendix B.

The basal period, or hour one (HRl), was analyzed for all

treatments to determine variation within animals between

treatment days in the rates of gastric fluid constituents.

Statistical analysis revealed no significant differences

between treatments, or within the entire control experiment

for volume, pH, hydrogen ion concentration, acid output or

osmolality. Pepsin concentration during hour 1 was not

significantly different between treatments. However, pepsin

concentration did vary markedly during the control



Gastric secretary rate tended to be higher and more

variable during the first hour (372 120 ml/15-min) than the

entire control treatment (310 78 ml/15-min, Figure 4.3).

During ranitidine administration, secretary rate was

significantly decreased from its baseline period only during

the last hour (P<0.05). During pentagastrin infusion, the

secretary rate was nearly doubled to 610 260 ml/15-min and

remained elevated for the entire infusion, but was

significantly decreased to baseline levels when ranitidine

was administered.

Gastric fluid pH was variable during the first hour

between treatments and averaged 1.77 0.46 versus 2.14

0.70 for the entire control treatment (Figure 4.4). The

control treatment had the most variation, and because of

this, between treatment comparisons had to have marked

changes to show statistical significance. Following

ranitidine administration, variation increased between

animals, especially at hour 2, where the pH ranged from 2.09-

5.24 and averaged 3.49 1.41. Analysis using the t-test

revealed a significant increase in pH for 2 hours following

ranitidine administration. When compared within each hour to

the other treatments, ranitidine increased pH above the

levels obtained only with pentagastrin at hours 2 and 5. This

was mainly due to an immediate increase following ranitidine

administration and the relatively long time before gastric pH

was significantly effected by pentagastrin. Gastric pH was

decreased to a minimum of 1.15 following pentagastrin

infusion which was significantly different from its own

control hour during the last 3 hours of infusion. This

decrease in pH was returned to control levels by ranitidine


Hydrogen Ion Concentration

Hydrogen ion concentration remained stable during the

control treatment (34 8 mEq/L, Figure 4.5). Ranitidine

produced an immediate decrease in H+ concentration to 17 6


mEq/L which was maintained for 2 hours during control

conditions. Pentagastrin infusion increased H+ concentration

maximally during hours 4 and 5 to 55 10 mEq/L which was

significantly different than control. Ranitidine

administration during pentagastrin infusion lowered H+

concentration to control levels for 3 hours.

Acid Output

The overall acid output during control was 46 13

AEq/kg/15-min, which was lower than hour 1 for all treatments

(62 22 AEq/kg/15-min, Figure 4.6). Ranitidine produced an

inhibition of acid output during all 4 hours when compared to

its own baseline collection (50-64% inhibition of control).

Pentagastrin produced a significant increase in acid output

for the duration of the experiment, but this increase was

inhibited (50%) for 2 hours by ranitidine administration.

Pepsin Concentration

Pepsin concentration was highly variable (110 34 Ag/ml)

and was the only variable where a significant difference was

demonstrated during the control treatment (HR2 < HR3, Figure

4.7). Ranitidine tended to decrease pepsin concentration for

one hour, but this was not a significant change from control.

Pentagastrin tended increase pepsin concentration for one

hour, but then decreased it during hours 3-5.

Pepsin Output

There was more variation in pepsin output than any other

variable (205 107 4g/kg/15-min, Figure 4.8). Ranitidine


tended to decrease pepsin output for one hour, but because of

the great amount of variation, a significant difference was

not demonstrated. Pentagastrin increased pepsin output

significantly during the first hour of infusion (HR2), after

which, pepsin output returned to control levels. Ranitidine

administration during pentagastrin infusion did not alter the

trends in pepsin output from pentagastrin infusion alone.


Gastric fluid osmolality was relatively stable (255 6

mosm/kg) and was less than plasma osmolality in all horses

(276 8 mosm/kg, Figure 4.9). A significant difference could

not be demonstrated for any treatment, although there was a

trend for gastric fluid osmolality to decrease following

rantidine administration, with or without pentagastrin

infusion. Pentagastrin tended to increase osmolality when

compared within hours.

Precision of Methods

Hydrogen ion and pepsin concentration and osmolality were

determined in replicate samples (Table 4.5). The coefficient

of variations for H+ titration and osmolality determinations

were acceptable, and for the H+ titration, indicated an

average error rate of about 3% for a given sample. The

coefficient of variation for the phenol method of pepsin

concentration determination was generally higher than the UV

method. In replicate samples, the coefficient of variation of

Legend for Figures 4.3-4.9.
These results are based on duplicate sets of 4
experimental treatments in 5 horses (protocol IV, see page
52). The control treatment involved a 5-h collection of
gastric fluid, during which a 0.9% saline drip was
administered at 200 ml/h. The second treatment was similar to
the control treatment, except ranitidine (0.5 mg/kg IV bolus)
was given at the beginning of hour 2. In the third treatment,
a pentagastrin infusion (6 Ag/kg/h) was started at the
beginning of hour 2. In the fourth treatment, a pentagastrin
infusion was started at hour 2, and then ranitidine (0.5
mg/kg IV) was given at the beginning of hour 3. Analysis of
variance was performed for two within-subject factors,
treatment (4 levels) and time (5 levels). Scheffe's multiple
comparisons procedure was used to determine significant
differences between treatment means at each hour with a P <
0.05 considered significant except for pH data where the t-
test was determined (see page 55). In the figures, a
significant difference for a treatment hour from its own
control (hour 1) is indicated by "a", while a significant
difference from the control treatment at the same hour is
indicated by "b".

M Ranitidine
=\\ Pentagastrin
=x JPentagastrin + Ranitidine

Mean+SEM (n=5)










1 2 3 4 5

Time (Hour)

Figure 4.3. Hourly secretion rate for protocol IV (see legend
on page 71 for treatment scheme; a= significantly different
from own control (hour 1) and b= significantly different from
control at same hour; P < 0.05).

C- Control
M Ranitidine
= Pentagastrin
Z Pentagastrin + Ranitidine











0 1 2 3 4 5

Time (Hour)

Figure 4.4. Hourly pH for protocol IV (see legend on page 71
for treatment scheme; a= significantly different from own
control (hour 1) and b= significantly different from control
at same hour; P < 0.05).

I= Control
M Ranitidine
X\N Pentagastrin
Pentagastrin + Ranitidine

0 1 2 3 4 5

Time (Hour)

Figure 4.5. Hourly acid concentration for protocol IV (see
legend on page 71 for treatment scheme; a= significantly
different from own control (hour 1) and b= significantly
different from control at same hour; P < 0.05).

Pentagastrin + Ranitidine









0 1 2 3 4 5

Time (Hour)

Figure 4.6. Hourly acid output for protocol IV (see legend on
page 71 for treatment scheme; a= significantly different
from own control (hour 1) and b= significantly different from
control at same hour; P < 0.05).

= -Control
M Ranitidine
= Pentagastrin
E Pentagastrin + Ranitidine








1 2 3 4 5

Time (Hour)

Figure 4.7. Hourly pepsin concentration for protocol IV
(see legend on page 71 for treatment scheme; a=
significantly different from own control (hour 1) and
b= significantly different from control at same hour; P
< 0.05).

M Ranitidine
= Pentagastrin
= Pentagastrin + Ranitidine








1 2 3 4 5

Time (Hour)

Figure 4.8. Hourly pepsin output for protocol IV (see legend
on page 71 for treatment scheme; a= significantly different
from own control (hour 1) and b= significantly different from
control at same hour; P < 0.05).


M Ranitidine
= Pentagastrin
SPentagastrin + Ranitidine


E 260
S250 T



230- \
0 1 2 3 4 5

Time (Hour)

Figure 4.9. Hourly osmolality for protocol IV (see legend on
page 71 for treatment scheme; a= significantly different
from own control (hour 1) and b= significantly different from
control at same hour; P < 0.05).


the lowest pepsin standard (5 Ag/ml) was 17% by the UV method

and 25% by the phenol method, indicating that both methods

had poor repeatability and precision at low pepsin

concentrations. The coefficient of variations of the other 3

standards ranged from 3.1 to 9.2 by either method.

Relationship Between the UV and Phenol Methods

The two methods were compared by two means. First, two

sets of 4 pepsin standards were measured in quadruplicate and

the regression lines for each method determined. The

coefficient of correlation between the 2 methods was 0.986

(P<0.000). The regression equations were:

(UV) = -0.052 + 1.087(PHENOL)

= 1.86 + 33.04* ABS

(PHENOL) = 0.060 + 0.895(UV)

= 0.375 + 35.45* ABS

where (UV,PHENOL) was the pepsin concentration in Ag/ml

estimated by either method. The slopes were close to 1

indicating good agreement between the 2 methods while the y-

intercepts were different.

Second, the regression lines from all standards of both

methods were:

(UV) = 3.04 + 33.86*ABS

(59 standards,replicated)

(PHENOL) = -1.74 + 53.55*ABS

(20 standards, replicated)


There is close agreement between the regression equations

from the UV method, using either 8 or 59 standards, while the

regression equation for the phenol method had a steeper slope

when more standards were analyzed.

Table 4.5. Statistical analysis of replicate samples.

Hydrogen ion concentration (n=5)
Sample 1 2 3 4 5 6
Mean(mEq/L) 25.3 42.9 62.8 74.8 92.2 50.1
STD 1.02 1.92 1.23 1.91 2.06 1.31
CV 3.98 4.48 1.97 2.56 2.24 2.62

Osmolality (n=5)
Sample 1 2 3 4
Mean(mosm/kg) 274.4 274.8 273.0 270.3
STD 1.14 1.48 1.77 1.50
CV 0.42 0.54 0.65 0.56

Pepsin standard absorbance
Sample 5 10 20 40
UV Method (n=4)
Mean(Ag/ml) 0.099 0.240 0.588 1.208
STD 0.017 0.019 0.015 0.038
CV 16.9 8.07 2.57 3.12
Phenol method (n=4)
Mean 0.113 0.285 0.638 1.264
STD 0.028 0.026 0.022 0.040
CV 25.0 9.22 3.50 7.57


Equine Gastric Fluid Secretion in Comparison to Other Species

A new procedure was developed to study gastric fluid

secretion in the young horse, a monogastric herbivore, and

revealed previously unknown parameters of equine gastric

physiology. The preparation was chosen to replicate normal in

vivo conditions and involved chronic, surgically prepared,

cannulated horses. Based upon the two-component hypothesis

for gastric fluid composition, the parietal component was

represented by acid output and the nonparietal component was

represented by pepsin output (99). Secretory rate and

composition were studied during basal (fasting) and

stimulated conditions. Stimulation of acid output was

achieved with a synthetic analog of gastrin rather than by

feeding, as the presence of food would have lowered the H+

concentration by dilution and neutralization 39). Histamine

has been shown to be important in the paracrine regulation of

parietal cell function so the effects of a H2-receptor

antagonist, ranitidine, were also studied in this model

(33,40,137,138). Vagal stimulation of acid secretion was not

studied, as previous attempts using 2-deoxyglucose were

unsuccessful in this preparation (see page 24). As well,


atropine was not selected as an antagonist of parietal cell

function due to the possibility of colonic impaction (12).

The following discussion will present these findings with

reference to gastric fluid secretion in other species,

especially the human and rat. These two species are relevant

for several reasons: 1) gastric and duodenal ulceration can

develop spontaneously in humans and gastric acid and pepsin

are the primary damaging agents (103), 2) the rat is commonly

used as a model for gastric and duodenal ulceration (140),

and 3) gastric mucosal anatomy in the rat is very similar to

the horse, being divided nearly equally into non-glandular

and glandular regions (135).

Since younger horses develop spontaneous duodenal

ulceration, animals between the age of 6 and 12 months were

studied. Age influences the rate of gastric fluid secretion

and it is therefore possible that there may be further

differences between 2-4 month old foals and these

experimental horses (1,3,87,70,107). These differences may

relate more to the amount of parietal or chief cell

responsiveness to stimulation or inhibition. It is therefore

expected that younger horses would also secrete acid and

pepsin during basal conditions, respond to pentagastrin by

increasing acid output, and respond to H2-receptor

antagonists by decreasing acid output.

External factors which could potentially stimulate basal

gastric acid secretion were minimized. Basal acid secretion


in humans displays a circadian rhythm with a peak in the

early morning (39). Each experiment was therefore started in

the morning between 7:30 and 9 am to control for any similar

rhythm in the horse. A fast of 24 hours was sufficient to

empty the stomach of all but a minimal amount of chyme. This

finding was in contrast to a report that some chyme remains

in the equine stomach after a 48-hour fast (73).

Contamination of gastric juice with swallowed saliva could

not be prevented in this model. It is assumed that salivation

was minimal since Alexander reported that parotid salivation

only occurs during mastication (6). The presence of a cannula

was clearly a deviation from the normal gastric condition but

this technique was the only one available to accurately

quantitate gastric secretion. Nasogastric aspiration of

gastric contents would not be successful, as the tube often

becomes occluded with chyme and its presence in the pharynx

stimulates salivary secretion (personal experience).

The position of the cannula in the left ventral abdomen

allowed collection of most gastric fluid by gravity. Phenol

red recovery rates indicated that some gastric fluid

continued to bathe the antrum and duodenum. Complete

diversion of gastric acid from the antrum and proximal

duodenum would remove inhibition of gastrin release resulting

in an artificial increase in basal acid output (40).

The phenol red recovery rate was lower in horse 1,

possibly because of the small volume of solution given (200ml


rather than 500ml) and lack of a 1-min sample. This may have

caused a false decrease in phenol red recovery because,

although phenol red is commonly used as a non-absorbable

intestinal marker, it may be adsorbed to surface epithelial

cells (130). A larger volume of administered fluid would

appear to have a better recovery rate because a smaller

proportion is lost due to surface adsorption. In subsequent

trials, the 1-min sample contained most of the phenol red

solution. A mean recovery rate of 83% for 12 horses with this

technique compares favorably with recovery rates from other

species with gastric cannulas.

The volume of gastric fluid collected displayed the least

intra-animal variation of all the fluid variables and was

constantly formed even during inhibition of acid secretion

with a H2-receptor antagonist. The increase in flow rate

observed during pentagastrin infusion was sustained and did

not suggest fatigue of parietal secretary mechanisms.

Rantidine decreased flow rate during pentagastrin stimulation

indicating that pentagastrin stimulated primarily the

parietal component.

Gastric fluid production in adult humans is nearly 3

liters/day, while pancreatic juice and biliary and salivary

secretions contribute another 4 liters/day to the upper

gastrointestinal tract (103). In contrast, combined parotid,

biliary and pancreatic secretions have been estimated to be

about 30 liters/day in a 100 kg pony (9). Argenzio reported


13.6 liters/day from salivary and gastric secretions and 18.2

liters/day from pancreatic, biliary and upper small

intestinal secretions in a 175 kg pony (11). During basal

conditions, the horses in the present study had an average

body weight of 200 kg and gastric flow rates of 1.4 liters/h

which could be stimulated to 2.5 liters/h. If these

stimulated rates could be extrapolated to feeding conditions,

these horses might secrete up to 30 liters of gastric fluid

per day. A combination of profuse salivary and constant

gastric, pancreatic and biliary secretions requires active

resorption of a large volume of fluid.

The purpose of such a large volume of secretion may be

related to the feeding habits of the horse. Gastric microbial

digestion of cellulose or other plant polymers allows

utilization of carbohydrates which are not hydrolyzed by

digestive enzymes (11,73). Under normal conditions, horses

are continual grazers and the gastric microbes would receive

a steady supply of substrate which saturated with

bicarbonate-rich saliva (4,9). Equine gastric microbial

fermentation of glucose yields pyruvate which is mainly

converted to lactic acid rather than volatile fatty acids as

in the pig and ruminants (11). Lactate concentrations peaked

in the stomach 4 hours after a meal in the stomach and

rapidly decreased aborad to the proximal small intestine to

very low levels in the cecum and large intestine (8,13).

Optimal conditions might thus be maintained for the

continuous production of lactic acid at a rate which may

contribute to the animal's nutrition.

In ruminants, the stratified epithelium of the rumen is a

site of volatile fatty acid, and Na+ and Cl- absorption (11).

The isolated stratified epithelium of the equine stomach

appears to be impermeable to ions and volatile fatty acids,

while the pyloric mucosa can absorb and transport volatile

fatty acids when the lumen pH is buffered to 7.4 (13). Such a

pH probably occurs during feeding as intragastric pH was

about 5.6 two hours after a meal (11,13).

Equine gastric fluid pH did not reach the minimal levels

reported for many other species during basal and maximal acid

stimulation conditions (Table 5.1) (12,40). This suggests a

large nonparietal component serving to dilute and neutralize

acids (HCl, lactic) and helping to protect the mucosa.

Periodic upward fluctuations in gastric pH, and hence

downward fluctuations in H+ concentration, were observed

during the control experiments and appeared unrelated to any

external factors. Such fluctuations have also been observed

by Russian investigators in the horse (145). Periodic

alkaline fluctuations in gastric pH have also been reported

in humans and occur mainly during the night or during non-

meal periods (23,68). The episode frequency (10.6 in 24

hours) is similar to the rapid spiking activity frequency

(phase III) of the migrating myoelectrical complex of the

proximal duodenum reported in the human (90). The neuro-


hormonal interaction between gastric secretion and proximal

duodenal motility may be highly organized to allow

coordination of digestive processes. In dogs, pancreatic and

biliary secretion seems to be coordinated with the phase III

of the proximal duodenum (131). In humans, basal pancreatic

secretion also seems to vary with the phase of interdigestive

motility (phases II and III > I and IV) (24). Gastric acid

secretion may be increased during the gastric phase III

(148). In anesthetized cats, direct vagal stimulation

stimulates gastric bicarbonate secretion which precedes an

increase in gastric acid secretion (112). However, it is not

known whether fluctuations in equine pH reflect

duodenogastric reflux or decreased gastric acid secretion (or

enhanced gastric bicarbonate secretion) in phase with

myoelectrical activity.

Gastric acid concentration is determined by: 1) the rate

at which it is secreted (acid output), 2) neutralization and

dilution by non-parietal secretions, and 3) back-diffusion

into the mucosa (probably minimal)(39,68). Acid concentration

is determined by titration to pH 7.4 with a standard base

(NaOH). Hydrochloric acid is probably the only acid substance

present during fasting conditions since lactic acid is

produced by the action of microbes on food and carbonic acid

would be rapidly converted to CO2 and water.

Gastric acid secretion has been studied primarily in the

human, dog, rat and rabbit (40,128,138). Significant species


differences exist in the rates of basal and stimulated acid

secretion (Table 5.1). Continuous basal secretion of both

gastric acid and pepsin, occurs in the human, monkey, pig,

chicken, rodent and herbivore (including the horse), but is

minimal in the dog and cat (72). Basal acid concentration in

the horse is similar to the human and pig (Table 5.1).

Basal acid output is high in the rat and rabbit (20 and

25% of maximal acid output, respectively), and very low in

the dog (1%) (40). Humans are intermediate in their basal

acid output, which is approximately 10% of maximal. Basal

acid output in the horse was 33% of the maximal induced by

pentagastrin stimulation. Basal acid output varies

considerably day-to-day in humans and from person to person

(coefficient of variation 80%) (48,150). Basal acid output in

the horse was also continuously variable between 15-minute

collections and between horses (coefficient of variation


The main purpose of performing a stimulation test is to

provide an estimate of the functional capacity of the stomach

(60,96). The choice of stimulant is based on preference and

absence of side-effects. The only side-effect noted in these

8 horses was in one that developed profuse salivation. The

route of administration is also by preference, but

intravenous infusion allows prompt termination of the drug if

any side-effects occur.


Step-dose techniques have been validated in the dog for

both histamine and pentagastrin (78). The calculated maximal

acid output could be linearly related to either single-dose

or step-dose techniques, but the full curves could not be

easily compared between single and step doses because of

"fade" or delayed responses. Pentagastrin, given by step-dose

technique, produced a similar maximal acid output at 3 and 6

gg/kg/h in 2 horses. The dose of 6 gg/kg/h was selected as

producing maximal stimulation because doubling the infusion

rate produced a decline in acid secretion. The dose used in

human secretary studies is 6 jg/kg given by subcutaneous

injection (48,103). In studies performed in one laboratory,

dogs and rats were shown to require a pentagastrin infusion

rate of 9 Ag/kg/h for maximal response, while rabbits

required 30 Ag/kg/h (40).

Pentagastrin stimulated acid output in the horse by

increasing both flow rate and acid concentration. The

increase in flow rate expressed ml/kg/h was much greater in

the horse than in the human or dog, while acid concentration

never reached their maximal values (Table 5.1). When acid

output is expressed as a unit of body weight to normalize for

large differences in body weights, pentagastrin produced

similar responses in the dog, cat, pig and rat, and less than

in these species in the human, horse and primate (Table 5.2)