Chronic kidney disease in horses is a rare diagnosis, with one study finding a prevalence of 0.12% in a hospitalised population, increasing to 0.23% in horses over the age of 15 (Reed et al, 2018). This is far lower than in companion animals, where the overall prevalence is reported to be 0.9% in dogs and 1.6% in cats (Reed et al, 2018). However, the prevalence in horses is likely underestimated as many older horses presenting to the ambulatory veterinarian will be euthanased without a specific diagnosis or referral to a hospital.
Chronic kidney disease is a progressive, terminal condition. The aim is to detect abnormalities as early as possible in the disease process and, with appropriate management strategies, maintain an acceptable quality of life until the point at which euthanasia is required on humane grounds. The rate of progression cannot accurately be predicted, as survival following diagnosis can vary from several months to over 2 years (Schott, 2018).
Causes
Mammalian kidneys have considerable reserve capacity. Loss of concentrating ability does not typically occur until two thirds of nephron function has been lost, and nitrogenous waste accumulation does not occur until three-quarters of renal function is lost (Osborne and Polzin, 1983). This understanding is based on subtotal and total nephrectomies so, although valuable, these should not be considered to represent naturally occurring disease. Renal failure is often a gradual process, so adaptive, compensatory changes are likely to occur. In horses, unilateral nephrectomy demonstrates that they can cope with the loss of 50% of renal tissue and compensatory hypertrophy of the contralateral kidney is seen (Irwin and Howell, 1980; Tennant et al, 1981; Sullins et al, 1988; Trotter et al, 1984; Juzwiak et al, 1988). However, before there is clinical evidence of disease, significant damage to the kidneys has typically occurred, which can make establishing the initial cause impossible. Clinical signs may also be subtle and might not be detected until late on in the disease course.
A complicating factor when attempting to ascertain the cause of chronic kidney disease in horses, unlike people and companion animals, is that by the time a horse reaches its mid-late teens it has typically changed ownership several times, meaning that pertinent elements of its clinical history are unknown. For example, risk factors include previous illness accompanied by prolonged hypovolaemia and treatment with nephrotoxic drugs, both of which may not be known by the current owner (Schott, 2007).
Broadly speaking, the causes of chronic kidney disease can be split into a number of categories which are outlined below.
Developmental
Developmental anomalies have been found to account for 16% of all cases of chronic kidney disease (Schott et al, 1997) and any horse diagnosed under 10-years-old should be suspected of having a developmental cause (unless there is a known previous risk of acute renal failure from nephrotoxic drugs or hypovolaemia). More severe conditions are incompatible with life and affected foals typically do not survive the fetal period or more than a few days following birth. However, less severe or unilateral conditions do not present until the total functional renal tissue is less that 30–40%. These cases often present later in life, following an event that a horse with normal kidneys could cope with:
- Renal hypoplasia: A smaller than usual kidney that therefore has less functional tissue (Andrews et al, 1986; Gull et al, 2001).
- Renal dysplasia: Failure of development of the glomeruli and tubules resulting in heterogenous population of nephrons ranging from near normal to non-functional blind ending structures (Anderson et al, 1988; Brown et al, 1988; Ronen et al, 1993; Jones et al, 1994; Gull et al, 2001; Plummer, 2006; Naylor et al, 2009; Gilday et al, 2015).
- Polycystic kidney disease: This is an inherited condition where cystic structures develop within the kidney parenchyma, obliterating functional tissue. However, this has rarely been observed in horses (Chandler et al, 2003).
Glomerulonephritis
Glomerulonephritis is an immune-mediated condition, which is more commonly a histological condition than it is a cause of clinical disease (Banks and Henson, 1972), likely because of the reserve capacity of the kidney. One study found glomerulonephritis to account for 53% of all acquired cases of chronic kidney disease in horses (Schott et al, 1997). However, later work by the same author suggests that this is an overestimation, with the true figure more likely being 10% (Schott, 2018). The cause is typically thought to be high levels of circulating immune complexes that occur with prolonged infections (Reed et al, 2018). These complexes are then deposited in the glomerular basement membrane. Typical infectious agents include equine infectious anaemia virus (Banks et al, 1972), Streptococcus equi subspecies zooepidemicus (Divers et al, 1992a) and equi (Roberts and Kelly, 1982). Leptospira pomona has been associated with subacute glomerulonephritis (Divers et al, 1992b) but has not been linked to chronic kidney diease in the horse (Divers et al, 2019).
There may occasionally be cases of a true autoimmune disease against the basement membrane (Banks and Henson, 1972; McSloy et al, 2007), although this is far less common.
Chronic interstitial nephritis
Schott et al (1997) also found that 39% of acquired chronic kidney disease cases were a result of disease starting in the tubules and/or interstitium (Schott et al, 1997). Chronic interstitial nephritis can be the consequence of several different disease processes that result in interstitial inflammation and fibrosis:
- Acute tubular necrosis because of ischaemia or nephrotoxic compounds (Gough and McGovern, 2019). This is the most common cause of chronic interstitial nephritis, although the clinician will often not have this information in the clinical history. In humans, ischaemic or nephrotoxic injury in the developing kidney can result in decreased nephron number and renal failure later in life, following apparent recovery from the neonatal illness (Andreoli, 2004). Whether this is the case in horses is unknown, but seems possible.
- Pyelonephritis as a result of ascending lower urinary tract infection (Tennant et al, 1981; Held et al, 1986; Carrick and Pollitt, 1987).
- Bilateral obstructive disease, such as ureteroliths or nephroliths (Ehnen et al, 1990).
End stage kidney disease
As many as 8% of cases may be called end stage kidney disease (Schott et al, 1997). This is where the extent of the disease is so severe at the time of presentation that it is not possible to determine the inciting cause. Grossly the kidney is small, pale and has an adherent capsule. Histologically there is extensive fibrosis and glomerulosclerosis (Reed et al, 2018). A suggestion of the cause may be obtained from the clinical history, but as stated previously this is often not known.
Other
Amyloidosis can also cause chronic kidney disease. However, unlike in dogs and cattle, in horses the kidney is a rare site of amyloid deposition (Jakob, 1971) and widespread amyloid deposition, secondary to multiple myeloma, did not have histological renal lesions in horses (Kim et al, 2005).
Renal neoplasia does not typically cause chronic kidney disease as it is normally unilateral and there is sufficient renal reserve capacity to compensate for this. However, in cases where there is additional renal compromise, either from congenital abnormalities or previous insults, chronic kidney disease may develop.
Clinical signs
The most common presenting clinical sign is chronic weight loss; reported to occur in 84% of cases. Other signs include ventral oedema (43%) (Figure 1), polyuria-polydipsia (44%), anorexia, rough hair coat and lethargy (Reed et al, 2018) (Box 1). Many of these signs are often misinterpreted by owners as ‘old-age’ or confused with other diseases such as pituitary pars intermedia dysfunction. Decreased performance may be noted in athletes and stunted growth is observed with some congenital abnormalities. With advanced disease, the horse may have a characteristic odour because of urea in the sweat and uraemic halitosis. Marked dental tartar can occur on the canine and incisor teeth and gingivitis and oral ulceration is observed. Uraemic encephalopathy can occur in end stage disease, resulting in abnormal behaviours, head pressing and seizures (Frye et al, 2001).
Box 1.Summary of clinical signs
- Weight loss
- Polyuria and polydipsia
- Ventral oedema
- Anorexia
- Lethargy/decreased performance
- Rough hair coat
- Dental tartar and gingivitis
The pathophysiology of the clinical signs is multifactorial and interlinked. Much of the understanding is extrapolated from studies in other species. For example, weight loss is likely to occur as a result of a combination of anorexia and the catabolic state driven by uraemic toxins (Schott, 2007).
Anorexia
In people, uraemic anorexia has been linked to elevated levels of plasma and central nervous system factors involved in appetite regulation and satiety, such as cholecystokinine, glucagon and serotonin, leptin and insulin. Proinflammatory cytokines are also thought to have a direct suppressive effect on appetite centres (Bergström, 1999). Urea itself has been shown to induce anorexia in dogs (Krawiec, 1996) and same mechanisms are likely to act in horses, although little direct work has been done in this area. Appetite might also be reduced by oral and gastrointestinal ulceration, which is likely to occur because of excess circulation of urea and ammonia, associated with reduced renal clearance and the prolonged half-life of gastrin because of reduced renal elimination. This results in increased gastric acid secretion and gastric ulcers (Schott, 2007). These effects can be confounded if a low palatability diet is offered as part of management attempts.
Ventral oedema
The development of ventral oedema (Figure 1) can be appreciated by considering Starlings forces:
- Net fluid movement = vascular permeability coefficient (capillary hydrostatic pressure – interstitial hydrostatic pressure) – reflection co-efficient to proteins (capillary oncotic pressure – interstitial oncotic pressure)
In chronic kidney disease there are perturbations in:
- Capillary oncotic pressure: Typically, it is believed that there is reduced oncotic pressure because of hypoalbuminaemia. However, in humans with nephrotic syndrome the capillary oncotic pressure and the interstitial pressure have been shown to decrease in parallel so the gradient is unchanged (Siddall and Radhakrishnan, 2012).
- Vascular permeability: The permeability is generally thought to be increased (Siddall and Radhakrishnan, 2012). This is thought to occur because of the direct effect of ureamic toxins on endothelial cells (Schott, 2007), and may also involve oxidative damage.
- Capillary hydrostatic pressure: Increasing renal hypoxia because of renal fibrosis results in activation of the reninangiotensin-aldosterone system. Renin is released from the juxtaglomerular cells, resulting in increased conversion of angiotensinogen to angiotensin I. Angiotensin converting enzyme then catalyses the conversion to angiotensin II, which promotes sodium retention, vasoconstriction, thirst and aldosterone release from the adrenals, with the overall effect being increased blood pressure (Ames et al, 2019) (Figure 2).
- Overall the result is net fluid movement out of the vascular system into the interstitium, resulting in oedema that gravity pulls ventrally (Figure 1).
Polyuria-polydipsia
Increased tubular flow, decreased medullary hypertonicity and decreased responsiveness of collecting ducts to antidiuretic hormone results in polyuria with compensatory polydipsia (Reed et al, 2018). This clinical sign may go unnoticed in horses kept at pasture. Other, more common, causes of polyuria and polydipsia should be considered, including pituitary pars intermedia dys-function and psychogenic polydipsia.
Lethargy and poor athletic performance
This is likely related to a mild-moderate anaemia thought to be caused by decreased production of erythropoietin by the kidney and shortened erythrocyte survival time as a result of uraemic toxins (Reed et al, 2018). In humans and companion animals the administration of recombinant erythropoietin is beneficial in the management of chronic kidney disease, but this has not been assessed in horses and is unlikely to be, since repeated doses in racehorses have been associated with severe, potentially fatal anaemia, which is believed to be a result of anti-recombinant erythropoietin antibodies (Piercy et al, 1998). In horses, recombinant erythropoietin serves as a temporary measure to improve morbidity while a patient waits for renal transplantation. Low energy levels may also be associated with partial anorexia.
Neurological signs
Uraemic encephalopathy is an uncommon finding in horses with chronic kidney disease, being seen in only 1% of renal cases (Frye et al, 2001). The pathophysiological mechanisms associated with this are unclear, although reactive astrocytosis has been observed on histology (Frye et al, 2001). A vast array of chemicals have been proposed as the responsible neurotoxins in humans, including urea, indoxyl sulphate, lanthionine, guanidine compounds, indolic acid, phenols and carnitine (Perna et al, 2016). It is believed that the clinical signs are the result of an imbalance in the inhibitory and excitatory neurotransmitters, neuronal degeneration and vascular inflammation (Popkov et al, 2019; Olano et al, 2021). Oxidative stress is thought to result in endothelial dysfunction, myelin injury, and nitration of brain proteins, leading to more uraemic toxin production (Vaziri, 2004) and worsening uraemic encephalopathy.
Diagnosis
A diagnosis of chronic kidney disease is generally based on indicative history, typical clinical signs and laboratory work. Initial serum biochemistry will reveal azotaemia; urea >10 mmol/litre and creatinine >220 µmol/litre. Serum creatinine levels do not increase until three-quarters of renal function is lost and therefore not a sensitive indicator of renal disease (Osborne and Polzin, 1983). Furthermore, in cases with a low muscle mass (as many are), the normal baseline creatinine concentration is low; a significant increase in creatinine does not result in levels exceeding the reference range and may remain undetected. In acknowledgement of this, in human and companion animal medicine, sequential increases in serum creatinine concentrations, rather than absolute concentrations alone, are considered in the classification systems (International Renal Interest Society, 2016), but there are currently no equine-specific classification systems. Similarly, increases in serum urea concentration offer low sensitivity for the detection of renal damage (Osborne and Polzin, 1983). Other factors including muscle catabolism, gastrointestinal haemorrhage and high protein diets will also increase serum urea concentrations (Mouton and Holder, 2006). The urea to creatinine ratio is typically >0.05 mmol/litre:µmol/litre; with higher ratios than acute disease (Schott, 2018).
To distinguish the observed azotaemia from pre-renal causes, a urine-specific gravity should be determined to assess renal concentrating ability. Urinalysis can be performed on free-catch samples. In chronic kidney disease, urine is isothenuric (1.008–1.014) as a result of damage to the renal tubules or interstitium, resulting in an inability to concentrate or dilute urine. As with azotaemia, concentrating ability is not a sensitive indicator of renal damage as a significant loss of nephrons can occur prior to loss of renal concentrating ability. When interpreting urine-specific gravity results it should be noted that heavy proteinuria can give spuriously high results (up to 1.020). In order to concentrate urine the kidney must have a hypertonic medullary interstitium and anti-diuretic hormone must be both produced and responded to. Other measures of urine concentrating ability have been described, including urine to serum creatinine ratios. Ratios exceeding 50:1 are considered consistent with concentrated urine (Grossman et al, 1982).
Blood samples may also identify a non-regenerative anaemia, hypoalbuminaemia and electrolyte disturbances. Although there is no specific pattern of electrolytes consistently seen in all horses with chronic kidney disease, the following derangements occur most commonly; hypercalcaemia, hyponatraemia, hyperkalaemia, hypophosphataemia and hypochloraemia (Schott et al, 1997). This may be the direct result of tubular damage reducing absorption or secretion of electrolytes, or alternatively, may occur secondary to other changes such as acid-base status or comorbidity.
The magnitude of hypercalcaemia is dependent on diet (Tennant et al, 1981). Unlike other species, this is not linked to parathyroid hormone, but to the kidneys' normal role in calcium excretion. Excess calcium absorbed from the gut is excreted via the kidney as predominantly calcium carbonate and a reduction in renal function results in increased serum calcium concentrations. This effect is exacerbated by hypoalbuminaemia, resulting in increased ionised calcium levels (Gratwick, 2020).
Sodium and chloride are freely filtered at the glomerulus and freely reabsorbed in the tubules of the healthy kidney. With significant tubular damage, reabsorption is reduced and hyponatraemia and hypochloraemia ensue.
In horses, hyperkalaemia appears to be more frequent (Fielding, 2015). In the healthy kidney, reduced glomerular filtration rate leads to increased aldosterone production, with a subsequent decrease in tubular potassium reabsorption and an increase potassium excretion (Khonsary, 2017). However, when glomerular filtration rate is reduced below a certain threshold, potassium excretion is reduced by an alternative mechanism and circulating potassium levels increase (Alcázar Arroyo, 2008). Blood pH and the concentration of other electrolytes can also affect potassium concentration (Aronson and Giebisch, 2011; Fielding, 2015) and the high potassium in an equine diet may also be significant.
The cause of hypophosphataemia is unclear, although postulated to relate to reduced feed intake (Gratwick, 2020).
Acid–base balance is likely normal through the majority of the disease but can see a metabolic acidosis in end-stage disease.
Markers of tubular dysfunction include measuring fractional clearance of electrolytes in the urine, compared with their levels in the plasma and measuring urinary γ-glutamyltransferase activity. However, both these parameters are typically considered more useful in cases of acute rather than chronic renal disease (Schott, 2007; Gratwick, 2020).
Reagent strips may show proteinuria, but to get an accurate measure urine protein concentration should be measured. Care should also be taken not to be misled by a 1+ protein on reagent strip, which may occur as a result of alkaline urine (Box 2).
Box 2.Blood and urinalysis
- Azotaemia
- Isothenuria
- Hypercalcaemia
- Hyponatraemia
- Hyperkalaemia
- Hypophosphataemia
- Hypochloraemia
Palpation per rectum may reveal a small, irregular left kidney, but the limited amount of kidney palpable on rectal examination may not allow this to be fully observed. In the case of obstructive disease, enlarged ureters in the retroperitoneal space may be palpated.
Transabdominal and transrectal renal ultrasonography can be useful and is relatively straightforward to perform (Box 3; Figure 3). Transrectal examination may be limited to the left kidney, depending on the size of the horse. When the renal cortex appears hyperechoic relative to the spleen and liver, or where there is poor differentiation between the cortex and medulla, chronic kidney disease may be suspected. However, increased cortical echogenicity is non-specific and the degree does not correlate with the severity of disease. Small irregularly shaped kidneys are also consistent with chronic kidney disease (or renal hypoplasia). The kidney of an adult horse is considered small when the long axis measures less than 10 cm (Matthews and Toal, 1996).
Box 3.Transabdominal ultrasound
- Prepare the scan site, clip if necessary and liberally apply contact medium
- 5 MHz transducer
Right kidney:
- Imaged in the 15-17th intercostal spaces
- At or slightly above a line connecting the ilium to the point of the shoulder
- 15 cm depth
Left kidney:
- Imaged in the 16-17th intercostal spaces
- On the ilium-shoulder line
- 20 cm depth
- Tip: The kidneys are always more ventral than you expect
Scintigraphic imaging of the kidneys provides information on renal vascular perfusion and excretion patterns, shape, size, location, patency and size of the collecting system and presence of space-occupying lesions (Schott et al, 1993). Although, this is infrequently performed and rarely gives sufficient detail (Figure 4).
The late stage at which these horses typically present means that distinguishing the original disease process may not be possible. Biopsy can also be performed (Gough and McGovern, 2019), but this rarely provides information that alters the treatment course and can carry significant risk, with complications being reported in 11% of cases and presentations including haemorrhage and colic (Tyner et al, 2011) Routine biopsy is currently not recommended.
There is a need for earlier recognition of renal disease. Serum symmetric dimethylarginine (SDMA) has been proposed as such a marker. It is produced in all nucleated cells, released into the circulation during protein degradation and excreted by the kidneys. The serum concentration of SDMA is proportional to glomerular filtration rate and it has been shown to be a more sensitive indicator of renal function than creatinine in companion animals (Hall et al, 2014). An equine serum SDMA test is now commercially available, with a suggested normal range of 0–14 µg/dl (Schott, 2018). Preliminary works suggests that this may be useful in the identification of acute kidney injury (Siwinska et al, 2020). However, the validity of this marker as an indicator of chronic equine renal disease is currently unknown.
A further molecule that may become a more sensitive indicator of renal disease is neutrophil gelatinase-associated lipocalin (NGAL), a glycoprotein that has been shown to be elevated in horses with renal damage and acute inflammation (van Galen et al, 2018). Before NGAL can be advocated for as a useful biomarker of renal disease, further work is required to differentiate increases associated with renal disease from systemic inflammation and establish normal ranges.
Treatment
It is not currently possible to reverse or even halt the progression of chronic kidney disease. A progressive decline in glomerular filtration rate and rise in creatinine will occur and treatment is therefore supportive.
At initial diagnosis intravenous fluid therapy can be beneficial in the case of an acute or chronic crisis. The aim of such treatment is to limit further loss of nephrons which would increase the rate of decline. However, these patients are at severe risk of fluid overload and the development of peripheral and pulmonary oedema. It is more appropriate to maintain fluid balance by provision of water ad libitum.
Attempts should be made to maintain acceptable body condition by feeding a palatable diet. High calcium and high protein feeds should be avoided (although not total protein restriction). If there are concerns that the protein content of the diet is too high, the blood urea nitrogen:creatinine ratio can be assessed and should be maintained at a target value of 10:1 (Schott, 2007). Good quality grass or grass hay is ideal. Fat supplementation can be a useful source of calories, although high levels of oil can be unpalatable. Ultimately, eating something is better than nothing so suboptimal feedstuffs should be fed where the inappetent horse declines the most appropriate options.
Pharmacological appetite stimulants have been explored with little success (Clark and Becht, 1987). Historically, in other species including cats, anabolic steroids were used as appetite stimulants. However, these are associated with a myriad of complications including electrolyte derangements and hepatotoxicity and are no longer recommended for use (Agnew and Korman, 2014). Mirtazapine, a serotonin antagonist, is commonly used as an appetite stimulant in cats with chronic kidney disease (Quimby and Lunn, 2013) and pharmacokinetic studies have demonstrated suitability for oral administration in horses (Giorgi et al, 2012), although the safety and efficacy of this has yet to be studied.
The correction of electrolyte abnormalities is controversial. Salt supplementation to address hyponatraemia and hypochloraemia would appear logical, but excess supplementation may exacerbate hypertension, with one study in cats demonstrating that supplementation with sodium chloride resulted in progression of renal disease, persisting even after salt supplementation was discontinued (Kirk, 2002). Salt supplementation should be avoided in horses at this time.
Antioxidants are theoretically useful in chronic kidney disease because of the progressive nature of the disease and the inflammatory processes. Their clinical usefulness is yet to be proved, however they are unlikely to cause harm.
Experimental evidence suggests that dietary supplementation with omega-3 fatty acids may slow progression of renal disease in humans and companion animals (Brown et al, 1998; Fassett et al, 2010). Several studies in humans look at the effect of omega-3 supplementation (such as inflammatory markers and vascular function) and it has been shown to increase quality of life scores in people undergoing repeated haemodialysis (Moeinzadeh et al, 2016), but to authors' knowledge, none have looked at survival. To date, there are no studies assessing the same benefits in horses but supplementation of the diet seems reasonable. Fresh grass is an excellent source of omega-3 and many commercial supplements exist. The dose rate in horses with renal disease is unknown, although supplementation with 1.5–3.0g of omega-3 is thought to be of benefit in horses with chronic lower airway inflammation (Nogradi et al, 2015).
Oedema should generally be tolerated. Diuretics should not be used to try and remove it, as this it risks further perturbations in electrolytes.
Controlling hypertension and reducing proteinuria is the mainstay of treatment in human medicine. Angiotensin converting enzyme-inhibitors may have benefits but are yet to be studied in horses. Benazepril (0.5 mg/kg per os twice daily) has been shown to be effective in healthy horses, although further work is required before their role can be advocated in horses with chronic kidney disease (Schott, 2018).
The use of steroidal and non-steroidal anti-inflammatory drugs to attenuate ongoing renal inflammation seems reasonable, but their blockade on renal prostaglandins limits renal vasodilation which can be detrimental. Currently their use is only advised in cases of glomerulonephritis with urine protein:urine creatinine >2:1 (Schott, 2007).
Specific treatments include immunosuppressive medications for glomerulonephritis. Cyclosporine has been explored in dogs but has not been shown to be beneficial (Vaden et al, 1995). It is likely that these interventions are being made too late in the disease process. It may be that advances in diagnostics allowing the earlier identification of renal disease could optimise the role of these drugs, at an earlier stage.
Ultimately in other species, renal transplantation is the treatment of choice, but this is not available for horses and is unlikely to become so. Maintaining the animal in the short term with peritoneal or haemodialysis is an option in theory, but without the transplantation end-point serious consideration is needed as to what the clinician is trying to achieve.
Prognosis
Ultimately, there is no cure available for chronic kidney disease and veterinarians currently have no ability to halt progression. The short-term prognosis is reasonable (from several months to a couple of years). The level of azotaemia is the greatest determinant of prognosis. The horse remaining appetent and maintaining body condition is the best outward sign. The speed of decline is impossible to predict at the outset.
Conclusion
Chronic kidney disease is typically a condition that has become advanced by the time of diagnosis. There is a need for more sensitive markers of renal damage or dysfunction so that this disease can be detected earlier in its course. SDMA and NGAL are potential biomarkers that are being investigated. Currently there are no options to halt the progression of disease, such that once the diagnosis has been made treatment is supportive only. However, the short term prognosis is reasonable so long as the horse remains appetent and maintains body condition.
KEY POINTS
- Chronic kidney disease is likely an underdiagnosed condition.
- Renal disease is likely advanced at the time of diagnosis.
- Biomarkers such as serum symmetric dimethylarginine and neutrophil gelatinase-associated lipocalin may prove useful for diagnosing of chronic kidney disease in the future.
- There is no treatment option to halt progression of disease
- The short-term prognosis is reasonable in horses that remain appetent and maintain body condition.