19 Jan 2015
Andy Durham
Job Title
ANDY DURHAM BSc, BVSc, CertEP, DEIM, DipECEIM, MRCVS explains the four general classes of diarrhoea in adult horses before advising on the most useful investigations to inform a choice of treatment
DIARRHOEA is a common problem in adult horses and varies in spectrum from the innocuous to the acutely life-threatening.
The study of diarrhoea in horses is impaired by a generally low diagnostic rate in clinical cases. Specific causes were found antemortem in fewer than 25 per cent of cases in two series of diarrhoea cases reported in adult horses and in 45 per cent of foals in another study.
The majority of cases remain idiopathic, although recent diagnostic aids have improved the diagnostic rate and are well worth employing.
There are four general classes of diarrhoea in adult horses that present in practice and it is helpful to categorise cases prior to deciding the diagnostic and therapeutic approach. Table 1 offers a brief outline of these four classes followed by more detail of diagnostic methods and treatment.
Horses are encountered that produce fairly normally formed faeces along with a separate “water phase” (Figure 1). Although frequently innocuous for the horse, this does bother owners considerably.
The condition remains well recognised, but somewhat mysterious, although the cause of many such cases may relate to fibre digestion in the colon and the limited capacity for water binding by poor quality or poorly masticated fibrous digesta.
Improving dietary fibre quality (for example, beet pulp, alfalfa) and mastication (dental care) can aid faecal material to hold water and improve the clinical signs in many instances.
Most commonly seen in equine practice are cases of mild diarrhoea in horses that are systemically well. Blood tests run on such cases return satisfactory haematological and serum biochemical results (proteins and inflammatory markers) and the affected horses are not significantly threatened by the effects of the diarrhoea.
Nevertheless, the condition probably warrants investigation and treatment. Aetiologic diagnosis of such cases has extremely low success and the cause is generally speculative. Occasional cases with evidence of parasitism or bacterial enteropathogens might be encountered, but most are often anecdotally attributable to dietary changes/intolerances, sand ingestion and so on. Treatment often involves dietary manipulation and, possibly, empirical probiotics.
Protein-losing enteropathy cases often present as weight loss rather than diarrhoea, although clearly this depends on cause and site affected. The diagnostic rate is high in this subgroup, as long as diagnostics (which can sometimes be invasive and expensive) are permitted by the owner.
Probably the most common cause and presentation is weight loss and diarrhoea in young horses in the winter, often a few weeks after deworming, a picture typical of en masse larval cyathostome emergence.
Various forms of inflammatory bowel disease (IBD) are poorly defined in horses, but nevertheless well recognised. Weight loss and intermittent colic are more commonly seen than diarrhoea in this author’s experience of equine IBD due to the predisposition of the disease to the small bowel.
Other causes of subacute-chronic protein losses that may (or may not) be associated with diarrhoea, and that are important to consider, include NSAID toxicosis, sand enteropathy and lymphoma.
In acute severe colitis, cases of diarrhoea are presented in which fluid, electrolyte, acid-base and protein disturbances along with endotoxaemia are acutely life-threatening. Such cases may progress rapidly and show signs of mild to severe systemic inflammation and shock when first presented (tachycardia, tachypnoea, discoloured membranes, pyrexia/hypothermia, cold extremities and others).
It is vital such cases are hospitalised early to maximise chances of survival. The diagnostic rate in such cases is reasonable, although many cases will have no aetiologic diagnosis.
Probable causes include post-antimicrobial diarrhoea, clostridial enterocolitis, carbohydrate overload, salmonellosis and cyathostominosis, although many of the previously mentioned milder forms of diarrhoea have the capacity to progress and destabilise the colonic floral balance, resulting in more severe diarrhoea.
High red cell indices (red blood cells, haemoglobin, PCV) might offer some guidance regarding hypovolaemia, but should not be relied upon. Anaemia is uncommonly encountered alongside diarrhoea, although if present, could reflect problems such as severe parasitism or lymphoma.
The leukogram is a very important part of the clinicopathologic investigation of diarrhoea. Neutrophilic leukocytosis may be encountered as part of a mild to moderate systemic inflammatory response in some cases of protein-losing enteropathy.
Cyathostominosis is commonly associated with especially high neutrophil counts. Neutropaenia is commonly associated with acute to peracute colitis cases as a result of loss of neutrophils into the inflamed colon wall and also the effects of endotoxaemia. Frequently, acute colitis cases will have circulating neutrophil counts of less than 1.0 × 109/L, which may persist for several days.
Accurate measurement of serum proteins is a very important part in the assessment and categorisation of diarrhoea cases in the horse. It is worth noting many benchtop analysers are loaded with entirely inappropriate reference intervals for serum proteins in horses and, therefore, great care should be taken when interpreting results from such machines.
The use of a refractometer to estimate total serum proteins is generally reasonably accurate and preferable to many small biochemical analysers. A normal adult horse will have a serum albumin concentration typically between 30g/L to 40g/L and a serum globulin concentration typically between 25g/L to 40g/L.
Many possible interacting and counteracting effects on serum proteins occur in association with diarrhoea in horses. Dehydration and hypovolaemia tend to increase the concentration of serum proteins, in contrast to mucosal inflammation during which significant losses of intravascular serum protein can occur.
These contrasting influences must both be considered when assessing an individual case. Thus, an apparently normal serum protein concentration and pattern in a horse that appears clinically dehydrated is likely to reflect hypoproteinaemia associated with enteric losses.
With subacute inflammatory protein-losing enteropathies, such as cyathostominosis, the typical serum protein pattern is one of hypoalbuminaemia (for example, 10g/L to 25g/L) and normal to high globulin concentration (for example, 25g/L to 50 g/L).
Although some still consider further fractionation of the total globulins using electrophoresis to have diagnostic usefulness, there is no evidence basis for this supposition and this test has an unacceptably low sensitivity and specificity for use in equine medicine.
The acute phase proteins, most commonly measured as plasma fibrinogen and serum amyloid A, are also an important part of the assessment and categorisation of diarrhoea cases. Initial assessment of serum proteins is a vital part of a fluid therapy plan.
Subacute protein-losing enteropathy cases will frequently achieve a pathophysiologic intravascular fluid balance based on remaining colloid and electrolytes. Intravenous isotonic fluids given to such cases can precipitate extensive oedema and imbalance of tissue fluid fluxes and this should always be preceded by effective colloid therapy to help retain fluids within the vasculature.
In this author’s experience, colloidal osmotic pressure (COP) below 15mmHg is potentially dangerous when isotonic fluids are administered without colloid.
COP can be estimated from the following formula:
COP (mmHg) = 0.55 × [albumin] + 0.25 × [globulin] – 4.4
Values of less than 20mmHg are normal and colloid therapy is indicated when COP is less than 15mmHg.
Oedema tends to be precipitated in acute protein loss when COP is less than 15mmHg, although more gradual chronic development of hypoproteinaemia tends to be tolerated until COP is less than 12mmHg.
Assessment of azotaemia should ideally accompany investigation of severe colitis cases, because dehydration in combination with other nephrotoxic insults, such as NSAIDs and aminoglycosides, has the potential to result in irreversible renal injury unless detected early and managed appropriately with fluid therapy and careful drug selection.
Lactate is a simple and useful reflection of hypovolaemia and endotoxaemia and should be measured wherever possible as part of the assessment of severe colitis cases. In addition to initial assessment, the response of plasma lactate concentration to the fluid therapy plan is a useful marker of success and prognosis.
Assessment of the nematode egg count in a faecal sample of diarrhoea cases is a frequent part of the investigation of relatively mild diarrhoea and/or subacute protein-losing enteropathy cases.
Although justifiable, it is important to realise a worm egg count is a poor reflector of parasitic burden in the horse. Inter-day variability of worm egg output is high and it is perfectly possible to have a large immature parasite burden in the absence of egg output.
It is frequently the case many cases of diarrhoea presented in practice have recently been dewormed by the owners as part of their empirical therapy for the horse’s problem. Therefore, the absence of faecal worm eggs is common, even if parasitism was the initial trigger for diarrhoea either as a result of adult or larval parasites.
The presence of dead or alive larvae in the faeces (Figure 2) supports a diagnosis of cyathostominosis although it could again simply reflect the recent empirical use of anthelmintics.
In certain geographic areas – such as Liphook Equine Hospital – sand enteropathy is commonly seen, associated with sparse summer grazing or grazing in the winter when grass is pulled up by its roots. Occasional cases are also seen in horses with an unusual appetite for soil (such as molehills). These may present as subacute protein-losing enteropathy and/ or various degrees of colic.
The presence of a significant amount of sand when a faecal sample is mixed with water and allowed to settle confirms the transit of sand via the gastrointestinal tract, although is arguably normal for any horse on sandy pasture.
Bacterial culture of diarrhoea samples is frequently disappointing. It is relatively unusual to encounter significant bacterial enteropathogens in cases of relatively mild diarrhoea in horses that are systemically well.
Indeed, even if an enteropathogen such as Salmonella was grown in such a case then its aetiological relevance could be doubted. With aerobic culture there are few unequivocal enteropathogens. Growth of a Salmonella species or Aeromonas species is frequently relevant. There is some evidence of potential enteropathogenicity of bacteria such as Enterococcus and Campylobacter, although more work needs doing in this area.
Anaerobic culture may reveal potential clostridial enteropathogens. However, the presence of Clostridium perfringens and, to a lesser extent C difficile, may not necessarily mean these potential enteropathogens are the cause of disease because some benign non-toxin-producing strains may occasionally be encountered.
Enzyme immunoassays are available for detection of C difficile toxins (toxins A/B), although the potentially relevant C difficile binary toxin cannot be easily tested. Nevertheless, presence of C difficile toxins A and/or B are likely to be aetiologically relevant when detected. Although C difficile colitis may occur as a primary event, it is frequently associated with post-antimicrobial diarrhoeas. Any horse developing diarrhoea in association with or soon after systemic antimicrobials should be tested for C difficile toxins. C perfringens may produce enterotoxins and be associated with moderate to severe colitis in horses.
Unfortunately, the presence by culture of C perfringens is not aetiologically meaningful. Additionally, the only readily available toxin test for C perfringens involves “C perfringens enterotoxin”. The specific association of this toxin with diarrhoea in horses has been questioned and other toxins such as beta-1 toxin might ideally be preferred, although not easily tested. Coronavirus has recently drawn attention as a possible equine enteropathogen that might occur as an outbreak including signs of diarrhoea, fever and possibly colic. Faecal ELISA tests may be used to identify the presence of the virus.
The main use of PCR in adult equine enteric disease is in detection of Salmonella in faecal samples. Although positive results do not necessarily mean the Salmonella is aetologically relevant, there are still implications of contagion with such a finding.
Examination of the colonic and caecal wall is an important aspect of investigation of diarrhoea cases, especially those with protein losses and/or systemic illness. Marked colon wall thickening (for example, 1cm mural thickness; Figure 3) should encourage more thorough investigation and therapy even when clinical signs do not appear initially worrying. Right dorsal colitis as a result of NSAID toxicity can easily be recognised by comparison of the right dorsal and ventral colons in the mid-right abdomen.
Diarrhoea may often be a frustrating disease diagnostically and inevitably many cases will be seen where no diagnosis is practically achievable. Nevertheless, considered selection of diagnostic tests alongside analytes useful in categorising the type and severity of the disease process will generally aid the clinical approach significantly.
Figure 1. A horse with normally formed faeces, but passing a separate water phase.
Figure 2. Cyathostome larvae on rectal sleeve.
Figure 3. Very thickened and folded colon wall in a case of cyathostominosis.
Table 1. Categorisation of diarrhoea cases with examples of causes
