INTRODUCTION

Equine gastric ulcer syndrome (EGUS) is a multifactorial problem that could involve one or several parts of the stomach of this species, including the terminal oesophagus, proximal (squamous mucosa) stomach, distal (glandular mucosa) stomach, and proximal duodenum ^1,2. The health of the equine stomach depends on the integrity of the mucosal areas (non-glandular and glandular), which is generated by protective mechanisms that counteract the effects of constant exposure to both, endogenous and exogenous gastroerosive agents ¹.

The caustic action of the gastric hydrochloric acid is a primary cause of EGUS ¹. However, as mentioned, EGUS is a multifactorial problem linked to the conditions of the animal (endogenous factors) and exogenous factors related to environmental conditions, such as the husbandry systems management and factors related with the wellness and healthiness of the animals with this condition. There is an extensive list of factors or direct causes involved in EGUS; these could include inherent factors such as gender, breed, age, and temperament. However, exogenous causes like training programs, level of competition, confinement level, concurrent medication (particularly, with non-steroidal anti-inflammatory drugs [NSAID]), inadequate feed management, and acid-fast bacteria producing metabolites, among others, are described in the literature as EGUS inducers ^1,4.

Volatile fatty acids (VFA) such as butyric, propionic, acetic and valeric acids are derived from carbohydrate fermentation processes directed by gastric bacteria. VFA have been found to produce "in vitro" lesions in the non-glandular mucosa of the equine stomach ^{(5, 6)}. However, in an early study it was observed that varying concentrations of VFA in stomach contents of horses fed diets based on hay or hay plus concentrate did not influence the appearance of gastric ulcers in these animals ⁷.

On the other hand, ulcerative lesions in the non-glandular mucosa are directly related to the action of gastric acid on this stomach area. In contrast, glandular lesions are more commonly a secondary cause of the systemic and local toxic effects of NSAID, particularly in horses medicated with phenylbutazone (PBZ) ⁸. The systemic side effects of PBZ are related to nonspecific inhibition of the cyclooxygenase (COX) enzyme, which is necessary for the synthesis of prostaglandins (PGE₂) and nitric oxide favoring protection and normal perfusion of the glandular mucosa. On the other hand, NSAID (e.g., PBZ) induce direct epithelial damage of glandular mucosa cells of the stomach ⁹.

Physiological protective factors related to the glandular mucosa are stomach blood perfusion, gastric bicarbonate production, and mucus secretion. These factors are mainly influenced by the local production of prostaglandin E2 (PGE₂). However, once COX-1 and 2 are inhibited non specifically by NSAID, such as PBZ, the PGE₂ concentration declines and glandular ulcers could appear ¹⁰.

Corn oil is a very good source of linoleic acid which is a precursor of arachidonic acid, which has the ability to increase PGE2 production ^10,11. However, there is scarce information on how this substance could affect the healing process in horses with EGUS and how this could affect the concentration of VFA in gastric juice in these animals. Therefore, we developed an equine model of PBZ-induced gastric ulceration 1) to investigate the influence of this substance on the VFA concentrations in the gastric juice in the animals; and 2) to correlate the gastroscopic findings with the VFA concentrations.

MATERIAL AND METHODS

Animals. Fifteen clinically healthy crossbred horses (11 not pregnant mares and 4 geldings) with an age range between 5 to 20 years and weight range of 270 - 465 kg were included. All the horses came from a free-pasture husbandry system. Two weeks before starting the experiment, the animals were conditioned by deworming and vaccination. The horses were allocated in individual stalls and fed Coastal Bermuda hay and water ad libitum. Furthermore, they received a concentrate supplementation with 13% protein and administered twice daily in the proportion of 1% of body weight.

Study design. This randomized complete block study included three specific times. At time 0, all the horses were evaluated in clinical laboratory and gastroscopic tests to obtain basal data. Time 1 was considered between the 1^st day and the 6^th day, and included the induction time. Time 2 included the time of evaluation of each treatment (treatment time), which was between the 7^th day and the 21^st day. This study was performed between May and July of 2012.

Experimental PBZ gastric ulcer induction groups. The 15 horses were randomly allotted into one of three groups, as a follows: in group I (control group), the animals only received 20 ml of water per os (PO), s.i.d, each 24 hours for 6 days; group II received a single PBZ dose of 13.2 mg/kg diluted in 20 mL water on the 6^th day of the start of the experiment; group III was medicated with a PBZ dose of 4.4 mg/kg, PO, s.i.d (always diluted in 20 mL of water) for 5 days and on the 6^th day received a final PBZ dose of 13.2 mg/kg, PO, diluted in 20 mL water. This induction protocol was designed to study the phenomena of gastroprotection and gastroadaptation the PBZ in the horse, proposed by Martinez Aranzales et al ¹².

Treatment groups. After 6 days of placebo (water) or PBZ gastric ulcer induction, the three groups of horses were treated as a follows: group I received 30 mg/kg sucralphate, PO, t.i.d, for 2 weeks; group II received 70 mL/100 kg of body weight corn oil, PO, twice a day, for 2 weeks; group III was treated with 90 mL/100 kg of body weight corn oil, PO, twice a day, for 2 weeks. The volume of corn oil used in each group, were extrapolated from work in the species ^10,13. The variables gender and age were blocked in this experiment.

Gastroscopy examination. Gastroscopy examination was performed on day 0, day 7 and day 21. After a solid fed fasting period of 12 hours and water deprivation time of 4 hours, the horses were sedated with detomidine (10 µg/kg/IV, Dormiun V, Agener União, Brazil) and a 12 mm -width- and 3 m -long- flexible videoendoscope (PortaScope^(r), 1800PVS, United State) was introduced into the stomach by nasogastric placement. The lesions observed were recorded and classified according to the scoring system of McAllister et al ¹⁴ (Table 1).

Determination of the concentration of VFA in gastric fluid. Prior to each endoscopy evaluation and with the help of a gastroscope, 5 mL gastric fluid (juice) were taken and deposited in glass flasks with 1 mL metaphosporic acid for later determination of the concentration of the VFA (acetic, propionic, butyric and lactic acids) by gas chromatography ⁷, through reference equipment Shimadzu (Shimadzu (Gas Chromatograph GC-17ª /auto inyector AOC - 20i Shimadzu).

Statistical analysis. Data were analyzed with the software SPSS 19.0 (IBM, SPSS Inc, USA). Lilliefors and Bartlett tests were used to assess normality and homoscedasticity in the studied variables. Therefore, data from butyric and acetic acids were transformed to log+4. The Tukey test and t test were used for comparison of means for acetic and butyric acid, respectively. Data from propionic and lactic acids did not present a normal distribution even after attempting several transformations. These data were compared by Friedman and Kruskal-Wallis tests. P <0.05 was considered statically significant for all the tests. The calculated power of this study was 0.8 (80%) using a sample size of 5 horses for experimental group and alpha error level of 5%.

This experimental study was approved by the local ethical committee for animal experimentation of the University Federal of Minas Gerais - Brazil, with protocol number 234/09. The study was conducted at the School of Veterinary Medicine at the University of Minas Gerais in Belo Horizonte, Brazil, located at latitude 19°55' S and longitude 43°56' W Gr, and an altitude of 832 m with kind of climate Cwa (according to Köpper classification), featuring winter predominantly dry summer rainy, according to the National Institute of Meteorology of Brazil.

RESULTS

All horses finished the period of the study without manifesting any clinical sign or metabolic alteration related to systemic illness.

Gastroscopy findings. At first gastroscopy assessment (zero day) no injuries were reported. Later on day seven post-induction, it was found that the protocol based in PBZ and even the effect of confinement produced lesions in the both non-glandular and glandular areas of the stomach. In general, these lesions presented a McAllister et al ¹⁴ degree for number and severity ranging from 1 to 3. Finally, at day 21 post-treatment it was found that all treatments (corn oil at two doses and sucralphate) produced healing of the injured the glandular area of the gastric mucosa. However, neither of the two corn oil treatments affected healing of the gastric ulcers located in the non-glandular area. Therefore, once the completed experimental phase, these animals were treated with proton-pump inhibitor (antacid) in traditional scheme for full recovery.

VFA concentrations in gastric fluid. The concentrations of VFA determined in the gastric contents in the two experimental phases are presented in Tables 2 and ³. The concentrations of acetic and butyric acids were highest in the gastric juice (2.32-3.80 and 1.64-2.39 mmol/L, respectively), where butyric acid represented most of the total concentration of VFA (Table 2). Propionic and lactic acids were present in lower concentrations (0.0015-0.12 and 0.05-0.61 mmol/L, respectively).

The baseline concentration (t0) of propionic acid showed significant differences between groups I and II. In the treatment phase, propionic acid concentration was significantly different in group II, between the second and third weeks. In the same phase, lactic acid concentration of group III was significantly different between the first and third weeks (Table 3). When the two experimental phases were compared, the concentration of propionic acid in group I showed a significant difference between the baseline and the final measurement. The same was also observed for lactic acid in groups II and III during the same period of time. Butyric and acetic acids did not present significant alterations in any groups and period of time evaluated during the study (Table 2).

DISCUSSION

EGUS is characterized by several predisposing factors that contribute to the high frequency of presentation of this gastric pathology in horses. Furthermore, the susceptibility of each of the areas of the stomach zones is also known. Gastric lesions are produced in horses by the action of NSAID and by situations with potential for generating stress, like confinement ¹⁵c. These two factors were considered in the present study for inducing gastric ulceration in the horses included. The protocol used in the present study generated ulcers in both, glandular and non-glandular mucosa. This situation allowed evaluation of the dynamics of the concentrations of the VFA under the effect of PBZ and participation of VFA in the healing of lesions during the administration of corn oil and sucralphate.

Horses in this study presented an absence of clinical signs and changes in their haemogram and biochemistry panel. These findings may be because these horses developed mild gastric ulcers with lower scores in number and severity¹⁴. However, the severity of an ulcer is not always correlated with the intensity of the clinical signs ¹⁶. This situation prevents determination of the clinical relevance of the degree of ulceration in some equine patients ^15,17. It is also important to consider the participation of the continued supply of food in maintaining gastric pH gradient, which results in temporary relief with little expression of clinical signs.

Total VFA concentration in the gastric contents of the horses in this study was higher compared with the concentration reported by Frank et al ¹³. In that study, corn oil and rice oil were added preventively to diminish gastric lesions associated with feed management. However, total VFA concentration of the present study was lower than those reported for diets based on concentrate and alfalfa hay ⁷. However, it is necessary to consider the nature of the present study, since the dynamics of VFA concentrations were measured in the presence of experimentally induced ulcers.

The differences found in the results of this study may derive from several factors that should be considered for their interpretation. In physiological conditions, the VFA concentration decreases markedly five hours after each meal ¹⁸. In the present study, sampling was performed after 12-14 hours of fasting. However, it is possible that the water restriction period was not enough to avoid its diluting effect on the gastric VFA concentration of the horses. On the other hand, the production of VFA is dependent on the type, quality, and quantity of food ⁷. This last factor varies widely in the different research studies. Furthermore, the bacterial flora adapted to the gastric environment is also involved in the fermentation processes; consequently, it may affect the VFA concentration ¹⁹.

Furthermore, it is likely that the total VFA concentration had also been changed by volatilization processes prior to the collection or by transformation to carbon dioxide, β-hydroxybutyrate, and acetoacetate, which normally occur with these acids ⁵. In addition, the animals in this study were fed a diet low in fermentable carbohydrates (approximately 2.5 kg of concentrated and distributed on average twice/day), unlike other studies where these types of carbohydrates were used ¹³.

Although, in the gastric ulcer induction period of this study no significant differences in the VFA concentrations were observed between the groups of horses, the VFA data showed slight alterations. However, as shown in the control group, the effect of confinement may have had a greater action on the gastric ulceration in the horses than PBZ. However, it is also necessary to consider that VFA have ulcerogenic effects in the equine stomach ^{(5, 7)} and are altered in horses with right dorsal colitis induced by PBZ ²⁰.

In vitro studies employing the Ussing chamber demonstrated the great ulcerogenic capacity of VFA. In this in vitro system, the concentration of each specific VFA is maintained in a constant fashion in the mucosa of the stomach ^5,6. However, this procedure excludes other natural components of the gastric environment that are involved in some way in its gastrolesive action. In contrast, an in vivo model includes all the components from the natural stomach environment, such as the motility, gastroduodenal reflux, interdigestive or basal condition, fluctuations in hydrochloric acid and gastric flora, amongst others, all of importance in the pathophysiology of EGUS. However, as in this study, some variables are characterized as being complicated to control or quantify.

The ulcerogenic effect of VFA depends on the pH in the gastric lumen. Specifically, at a pH lower than 4, VFA accentuate the alterations in the non-glandular mucosa area of the equine stomach, specifically in the area adjacent to the margo plicatus the lesser curvature ^{(5, 6)}. In the present study, the perpetuating injuries focused in the non-glandular mucosa area were possibly derived due to the no ionization of the VFA assed (acetic, propionic and butyric acids) and, the low pH of gastric contents (2.38-3.77) in the groups treated with corn oil, which the ulcers did not completely heal.

It is know that VFA modify the bioelectric properties of the gastric mucosa by decreasing sodium transport. This results in different levels of change in the cell membrane permeability, according to the type of acid ^5,6,21. Acetic acid showed the highest concentration (59%) in the gastric contents of the horses of this study. This result is similar to that of other studies ^5,6,7, but it has been reported that in horses treated with corn and rice oils, this VFA may reach up to 85% of gastric contents ¹³. However, acetic acid causes discrete lesions in the gastric mucosa.

Contrary to what was reported by Frank et al ¹³, the gastric contents of this study showed butyric acid concentrations of 40%. "In vitro" studies have shown that butyric acid has a more damaging effect on the non-glandular mucosa area ⁵. The higher butyric acid concentrations reported in this study may be related to the extensive ulceration observed in the non-glandular mucosa area of the horses in this study.

Propionic acid has a moderate gastrolesive effect ⁵. However, the concentration of this VFA in the gastric juices evaluated was lower (0.0015-0.12 mmol/L) than those reported in the literature. Therefore, the ulcerogenic role of propionic acid was possibly minimal in the animals in this study. The differences in the concentration of propionic acid between different studies are possibly associated with the aforementioned factors, such as feed manipulation and prolonged fasting periods.

There are several studies that describe abundant production of lactic acid in the equine stomach ¹⁹, especially immediately after consumption of food ¹⁸. In contrast, the concentrations of lactic acid reported in the present study were lower in comparison with those of other studies. The fasting time and type and feeding scheme may have influenced these results. In vitro studies have demonstrated that VFA has minimal ulcerogenic capacity in the non-glandular mucosa of the horse ²¹. The effect of lactic acid is dependent on the amount and time of exposure to the mucosa; therefore, it is possible to infer that this VFA plays a poor role in the generation of gastric ulcers in the horses of this study.

In conclusion, the corn oil and sucralphate treatments did not lead to differences in the concentration of acetic acid and butyric acid in this study. In contrast, these treatments produced different gastric juice concentrations of propionic acid and lactic acid. However, the small quantity and proportion of these VFA with respect to acetic acid and butyric acid, let us to think that propionic acid and lactic acid were no relevant VFA during the induction and treatment periods in the horses of this study. Conversely, the gastric acids with the highest concentrations, in particular butyric acid, possibly perpetuated the lesions in the non-glandular mucosa area, where the corn oil showed no therapeutic benefit. These results verify the harmful effect of some AGV on non-glandular mucosa, however it should be considered restorative action through induction of prostaglandins per treatment in the glandular mucosa.

Finally, VFA must be continuously measured over short periods of time to eliminate the effect of volatility and metabolic transformations of these substances in the equine gastric environment. Furthermore, it is necessary to bear in mind the dynamics of the phases of emptying of the stomach to reduce their impact on the results and to effectively compare new results with reported "in vitro" studies.