It is widely accepted that equine gastric glandular disease is a separate entity to equine squamous gastric disease (Rendle et al, 2018). The risk factors (Sykes et al, 2019) and response to treatment (Sykes et al, 2014; 2015a) of equine glandular gastric disease differ significantly from equine squamous gastric disease as a result of differences in the anatomy and pathophysiology of the two regions. Over the last 10–15 years, there has been increased identification and reported prevalence of equine glandular gastric disease. However, there is still a significant lack of knowledge around this condition; some of the reasons for this will be discussed in this article.
Prevalence and risk factors
The prevalence of equine glandular gastric disease has anecdotally increased over the last 15 years; in clinical and abattoir studies of various horse types worldwide, prevalence is between 47% and 65% (Begg and O'sullivan, 2003; Nieto et al, 2004; Luthersson et al, 2009a; 2009b; Martineau et al, 2009; Husted et al, 2010; Tamzali et al, 2011; Binns et al, 2016; Sykes et al, 2019).
Known risk factors are sparse and occasionally contradictory, but are very different to those for equine squamous gastric disease. Warmbloods are at increased risk compared with other horse types (Luthersson et al, 2009b; Mönki et al, 2016). In Thoroughbred racehorses, the horse's trainer was identified as a risk independent of other management factors (Sykes et al, 2019). Exercising for more than 4 days per week is a risk factor in racehorses (Sykes et al, 2019) and sports horses (Pedersen et al, 2015), whereas intensity of exercise was not. The more experienced a horse is at its discipline, the lower prevalence of glandular gastric disease (MacLeod et al, 2015; Pedersen et al, 2015), which may suggest work adaptation or differences in management of elite horses (Rendle et al, 2018). A study in endurance horses demonstrated a higher prevalence of glandular gastric disease in the competition season (Tamzali et al, 2011) which may relate to reduced gastric blood flow during exercise.
Horses with severe glandular gastric disease have increased cortisol concentrations in response to novel stimuli (Malmkvist et al, 2012), and in response to exogenous adrenocorticotropic hormone (Scheidegger et al, 2017), suggesting more sensitivity to stress. The association between equine glandular gastric disease and stereotypies is conflicting; no association was found in one study (Sykes et al, 2019), whereas an association with crib-biting was found in another (Scott et al, 2017). These differences may be explained by a high prevalence of both conditions in each population, particularly as stereotypies are a coping strategy in horses. It would be useful to evaluate disease prevalence in animals prevented from exhibiting stereotypical behaviour. The lower prevalence in more experienced polo ponies and showjumpers (MacLeod et al, 2015; Pedersen et al, 2015) may relate to adaptation to physiological stress. More experienced show and showjumping horses also have lower cortisol concentrations than less experienced horses (Covalesky et al, 1992; Cayado et al, 2006). It is challenging to know what is ‘stressful’ to an individual horse and, as such, changes to minimise stress should be tailored to an individual and ideally kept consistent.
No association has been documented between equine glandular gastric disease and infectious agents (Martineau et al, 2009; Husted et al, 2010), non-steroidal anti-inflammatory drug use in clinical cases (MacLeod et al, 2015; Pedersen et al, 2015; Mönki et al, 2016; Sykes et al, 2019), diet or lameness (Rendle et al, 2018; Sykes et al, 2019).
Pathophysiology
Lesions of the glandular mucosa are not ulcerative; instead, they are erosive and inflammatory in nature, consisting of a mixed inflammatory population (lymphocytes, plasmacytes and neutrophils) and as such, this condition fits the description of a glandular gastritis (Martineau et al, 2009; Husted et al, 2010; Crumpton et al, 2015).
The glandular mucosa is different to the squamous mucosa in that it is adapted to the highly acidic (pH 1–3) environment that it is usually bathed in (Merritt et al, 2003). It is likely that glandular gastric disease may result from a breakdown of the normal defence mechanisms that protect the mucosa (bicarbonate and gastric mucus, which is made up of glycoproteins, water, electrolytes, lipids and antibodies) (Hepburn, 2012).
The proposed pathophysiology of these lesions includes changes in blood flow with or without perpetuation of these lesions because of acid exposure (Sykes et al, 2015b), or as an extension of inflammatory bowel disease (Rendle et al, 2018). Reductions in blood flow may be secondary to ‘stress’, which influences gastrin production or be associated with exercise and feeding (Rendle et al, 2018). Muñoz-Prieto et al (2022) found that horses with glandular gastric disease had upregulated salivary proteins relating to immune activation compared with healthy horses, which may support the proposal that glandular gastric disease is a manifestation of inflammatory bowel disease.
Clinical signs
Clinical signs associated with equine glandular gastric disease are non-specific and overlap with those associated with equine squamous glandular disease. They include changes in temperament such as nervousness and aggression, changes in rideability including reduced willingness to work and reluctance to go forwards, unexplained weight loss (likely concurrently associated with reduced appetite or altered eating patterns), cutaneous hypersensitivity manifesting as flank-biting, resentment to girthing, leg aids or rugging, mild and/or recurrent abdominal pain (Rendle et al, 2018). Based on the current literature, changes in coat condition, stereotypical behaviour, bruxism or diarrhoea are unlikely to be associated with glandular gastric disease (Rendle et al, 2018).
Cutaneous hypersensitivity seems an implausible clinical sign of gastric disease. However, in other species, afferent pathways from the abdominal viscera and 6–9th thoracic spinal nerves are pooled such that those signals from the skin may be affected by input coming from the viscera and misinterpreted within the brain (de LaHunta et al, 2014).
Diagnosis
Gastroscopy remains the only method for diagnosis, with the majority of lesions located around the pylorus and pyloric antrum (Murray et al, 2001; Sykes et al, 2014; 2015a; 2019). The grading system does not reflect severity of disease, and until a better system is developed, describing lesions based on the ECEIM-ACVIM consensus statement is recommended (Sykes et al, 2015b). These descriptors include focal, multi-focal and diffuse; mild, moderate and severe; nodular, raised, flat or depressed; erythematous, haemorrhagic or fibrinosuppurative. A more recent consensus statement focused on equine glandular gastric disease suggested that flat, erythematous lesions were likely to heal more rapidly than those with a nodular, raised, fibrinosuppurative or haemorrhagic appearance (Gough et al, 2020). The agreement regarding these descriptors varies between fair to moderate (Tallon and Hewetson, 2021) and moderate to good (Pratt et al, 2022). This poor agreement and lack of reliable methods for grading severity of lesions makes it challenging to perform clinical and experimental studies on this disease, and also to make comparisons between studies. As such, defining and describing improvement, worsening and healing is more valid and valuable.
Differences in the gastric microbiota between healthy horses and those with glandular gastric disease have been documented (Paul et al, 2021; Voss et al, 2022), but it is currently uncertain whether these are cause or effect and how they can be used clinically.
Treatment
A number of treatment combinations have been proposed (Rendle et al, 2018): oral omeprazole (4 mg/kg every 24 hours) combined with sucralfate (12 mg/kg every 12 hours), oral misoprostol (5 μg/kg every 12 hours) with or without sucralfate (12 mg/kg every 12 hours); long-acting intramuscular omeprazole (4 mg/kg every 5–7 days), which is available in some locales, esomeprazole or therapeutic combinations for inflammatory bowel disease (including parenteral steroids).
Sucralfate has a range of effects in the stomach. These include creating a physical barrier that protects against acid, stimulating mucus secretion to block acid diffusion, inhibiting pepsin and bile acid secretion, promoting epithelialisation and increasing mucosal blood flow through increased production of prostaglandin E (Barton and Hallowell, 2023). Response to oral omeprazole and sucralfate therapy varies, with healing rates reported between 22% and 63% (Hepburn and Proudman, 2014; Varley et al, 2019); this difference likely relates to differing definitions of healing – one of many research challenges. When used, omeprazole should be administered on an empty stomach and the horse should not be fed for 30–60 minutes after administration (Rendle et al, 2018).
Misoprostal is a prostaglandin E analogue and should improve mucosal blood flow. It also suppresses acid production in the horse (Sangiah et al, 1989) and inhibits neutrophilic inflammation (Martin et al, 2017); Varley et al (2019) demonstrated a healing rate of 73%. Side effects are rare, but include mild, transient diarrhoea, mild abdominal pain and urticaria. Care must be taken in administration to pregnant mares as this drug could induce abortion, although there are some safety data to suggest it can be administered between 100 and 130 days gestation (Jacobson et al, 2013). Because of its abortigenic potential, this drug should not be dispensed to owners or caregivers who are pregnant or planning to be pregnant. There is no rationale for combining this drug with oral omeprazole (Rendle et al, 2018). There is anecdotal evidence and one published report (Barton and Hallowell, 2023) of development of squamous lesions while using this drug in conjunction with improvement in glandular healing. The author has also seen this phenomenon in a similar percentage of cases treated with injectable omeprazole and does not view this as a particular problem with either drug, as the squamous disease can easily and swiftly be treated, although further probing of management and nutrition should be encouraged. The aetiology for this is unknown, but may relate to a rebound acid effect or lack of identification of other risk factors (Barton and Hallowell, 2023).
Long-acting intramuscular omeprazole is more effective than oral formulations for acid suppression when pH is measured in the ventral portion of the stomach. Acid suppression is maintained for 4–7 days, so the injection should be administered at 5-day intervals (Sundra et al, 2024a). Healing rates of 64–75% have been reported (Sykes et al, 2017; Gough et al, 2022). Transient swelling at the injection site has been reported in <10% of cases, so it is recommended to be administered after warming, into the gluteal muscles.
Administration of glucocorticoids may be warranted in some groups of gastric disease patients where the disease may be an extension or manifestation of more generalised inflammatory bowel disease and have been reported to be effective (Martineau et al, 2009; Husted et al, 2010; Rendle et al, 2018). Initial administration of 1 mg/kg prednisolone orally every 24 hours or 0.05–0.1 mg/kg dexamethasone orally every 24 hours, which is then gradually tapered over 4–5 weeks, has been proposed (Rendle et al, 2018). Other recommendations include dietary simplification, where cereal proteins or alfalfa may play a role in the initiation or perpetuation of inflammatory bowel disease (Barton and Hallowell, 2023).
There is no evidence for administration of antibiotics, ranitidine, aloe vera, pectin-lecithin complexes, polysaccharides, kaolin, bismuth subsalicylate, sea buckthorn, acupuncture or homeopathy for the treatment of equine glandular gastric disease (Rendle et al, 2018). Based on these findings and the need for judicious use of antimicrobials, veterinarians should not be using antibiotics as a treatment option without proven infection based on culture and sensitivity. However, many of these studies are hampered by challenges with study design (experimental induction of lesions), low sample size, use of less ideal statistics and trying to compare nutraceuticals (which, by their definition, are not medicines) with pharmacological agents. Where squamous and glandular gastric disease occur concurrently, treatment should be aimed at the glandular disease, as any of the treatment for this would most likely resolve squamous lesions.
Expectations for healing and monitoring
Rates of healing for equine glandular disease are slow and unpredictable compared with those of squamous gastric disease. Raised, nodular and fibrinosuppurative lesions may take longer to heal than flat, haemorrhagic lesions, although further studies are required to evaluate these observations. In one study, only lesion severity correlated with healing (Pratt et al, 2023). Mucosal restitution can occur within 3–5 weeks, but may take several months to completely resolve, particularly where raised areas or nodules are visible (Rendle et al, 2018).
Evaluation should be performed every 6–8 weeks using gastroscopy until resolution has occurred, and only then should treatment be discontinued (Rendle et al, 2018). In some populations and with certain lesion types, examination before this time point (eg at 4weeks) may be warranted. There is no rationale for reducing the dose of the drugs, except for glucocorticoids (Barton and Hallowell, 2023).
Management of the refractory case
If there is no improvement or deterioration in lesion appearance at the re-examination, it is recommended to change to an alternative first-line treatment. If there is improvement, first-line treatment should be continued for a maximum of 3 months and alternative treatments considered at that point (Rendle et al, 2018). A small proportion of cases develop very large hyperplastic nodules that are at risk of obstructing pyloric outflow (personal communication). These cases may require transendoscopic removal using thermocautery or LASER.
Prevention of recurrence
Prevention is problematic. Based on known risk factors, horses should have 2 rest days per week, stress should be minimised (calm environments, minimal number of carers and the same equine field companions) and turn out maximised (unless that is deemed stressful) (Rendle, 2018). In the authors’ opinion, any potentially gastro-irritant supplements (such as magnesium sulphate) should be stopped. Corn oil (150–250 ml/day/500 kg) may be beneficial, as it decreases gastric acid output and increases prostaglandin E (Murray et al, 2001), and the use of pectin-lethicin as a mucosal protectant at 150 g every 12 hours (or feeding sugar beet pulp) may be beneficial (Cargile et al, 2004; Rendle, 2018).
Why has there been so little progress in the last 10 years?
Many researchers and groups have attempted to determine more information regarding this unique form of gastric disease in the last 10–15 years. Some had conceptions that the disease was like equine squamous gastric disease, which were quickly refuted, and a lot has been learned about what the risk factors are not, and what treatments do not work (Rendle et al, 2018; Sykes et al, 2019).
A great deal of information is missing, including the underlying pathology and pathophysiology. Two abstracts on findings with full thickness post-mortem biopsies (Crumpton et al, 2015) and transendoscopic glandular mucosal biopsies (Hepburn, 2012) present somewhat conflicting data – one that the lesions seen at the pylorus do not reflect the extent and degree of the underlying gastritis, and the other that the transendoscopic pyloric biopsies do correlate with assumed lesion severity.
The next challenge is regarding experimental studies using feed deprivation or phenylbutazone-induced ulcer models (Murray and Eichorn, 1996; Ricord et al, 2021). While these do make for controlled experiments, they do not reflect the disease seen in clinical patients. They are reliable models for the development of squamous gastric disease, but they do not consistently result in the development of glandular gastric disease lesions. These lesions may also spontaneously heal in a way that does not occur in natural disease.
There have also been studies which attempted to evaluate nutraceuticals as medicines (Huff et al, 2012; Bush et al, 2018). If these were medicines, they would be marketed and managed under different regulations. While it is acceptable to look at the effects of preventing recurrence of disease, it is entirely inappropriate to compare the healing effects of a nutraceutical with a medicine with known efficacy against glandular disease such as omeprazole and misoprostol. These types of study, often performed in a small number of clinical cases, can lead to the belief that these nutraceuticals do not have a role to play, which they may do if they were evaluated as having adjunctive roles in the management of the disease.
There are also challenges regarding the categorisation of the type and severity of disease. What can be visualised at the pylorus using gastroscopy may not represent the extent of the gastric disease as repeatability of assessment of perceived severity is poor and it is unknown which type of lesions are severe outside of anecdotal evidence. This provides challenges when categorising disease for statistical evaluation, and many studies with multiple categories result in very small numbers in each category. This makes it challenging to identify true change and increases the likelihood of type 2 errors. Further work is needed on which descriptors are most valuable regarding underlying pathophysiology, likely response to specific therapies and likely speed of response to therapy.
Studies continue to use a grading system that has been widely accepted to be unfit for purpose or have started using other, nonvalidated systems (Hwang et al, 2022; Sundra et al, 2024b), which include subjective grades of mild, moderate and severe. The previous grading system for equine gastric ulcer syndrome also does not represent a ‘continuum’ of more to less severe lesions (Rendle et al, 2018; Lo Fuedo et al, 2022), which is how it was often statistically evaluated. The more categories that are evaluated statistically can impact on result validity and type 2 errors, depending on the number of horses included. Analysis of grading systems using parametric tests and reporting of mean and standard deviation is inappropriate and means that results cannot be accurately interpreted (Huff et al, 2012).
What should be used based on current information?
Although basic, using improvement, healing and worsening can be very powerful (when analysed using a non-parametric test) as long as there are clearly defined criteria for improvement and healing. The main limitation of this approach currently is differing definitions in different publications – for example, some have defined healing as mucosal restitution (Hepburn and Proudman, 2014) and others define healing as the return of normal tissue (Varley et al, 2019), which results in findings being incomparable.
To increase veterinary knowledge of this disease, there are several aspects which must be addressed. Large-scale post-mortem studies with histopathology are required to allow better understanding of the likely significance of what is seen at the pylorus, whether different lesion types are associated with a different underlying histopathology and which descriptors correlate best with severity or likely speed of healing. Large, multi-centre studies are required to observe the impact of glandular disease on clinical signs, while ruling out or managing concurrent clinical signs that could produce similar signs, such as lameness and back pain. Consistent definitions of healing and improvement of lesions to directly compare treatments must be decided. Nutraceuticals should be evaluated as adjuncts and not as medicines, and appropriate statistics should be used to increase confidence in the reported results.
Conclusions
This article provides an update on what is and is not known regarding the pathophysiology, significance and severity of lesions, and likely response to therapy. While knowledge has significantly increased, this disease is not fully understood.