Early Diagnosis of Chronic Kidney Disease

Adapted from the IRIS education article by Harriet M. Syme (London, UK), an update of an earlier article by AM van Dongen (Utrecht, Netherlands) and R Heiene (Oslo, Norway), 2013.

Introduction

In clinical practice, chronic kidney disease (CKD) in dogs and cats is unfortunately usually identified quite late in the disease process, often once the patient already has clinical signs. This limits the benefit of treatment, which in many cases can be expected to slow progression rather than restore kidney function, and it makes the underlying cause difficult to identify. Developing better methods for the early detection of CKD is therefore a goal in both veterinary and human nephrology.

It is often assumed that we cannot detect CKD until a dog or cat has lost more than three-quarters of its nephrons and developed azotemia because the conventional tests, urea and creatinine, are insensitive indicators of glomerular filtration rate (GFR). It is true that the relationship between these indirect markers and GFR is curvilinear, so that when kidney function is near normal, GFR has to change substantially before the markers rise noticeably. But the ability of the remaining nephrons to enlarge and increase their own filtration (hypertrophy and hyperfiltration) is often overlooked. For example, if a patient loses 50% of its kidney function through removal of one kidney, GFR will have nearly returned to its pre-surgical value within a few weeks. Therefore, even with direct measurement of GFR (see GFR in practice), it can be difficult to detect CKD in its earliest stages, particularly if the disease is tubulointerstitial.

Identifying Patients With CKD but Normal GFR (IRIS Stage 1)

By definition, dogs and cats with Stage 1 CKD have normal or near-normal GFR. This means that many of the clinical signs classically associated with CKD are absent, and many, but not all, patients with Stage 1 disease have no signs at all.

Some patients are diagnosed with IRIS Stage 1 CKD through screening because of a family or breed history of a kidney disorder for which a specific diagnostic test is available, such as polycystic kidney disease in Persian cats or hereditary nephritis in cocker spaniels. In these patients, GFR is lost progressively as the disease advances, and they will eventually develop azotemia.

Some patients are recognized because a reduced ability to concentrate urine shows up as increased urination and thirst (polyuria and polydipsia). When other causes have been ruled out, CKD should be considered.

Other dogs and cats are diagnosed with IRIS Stage 1 CKD because of structural kidney lesions, which are often discovered incidentally during an abdominal ultrasound for an unrelated problem, for example kidney infarcts, kidney cysts, or poor corticomedullary definition. In these patients, it is often difficult to judge how likely the disease is to progress and whether any treatment for CKD is needed. Patients may also be diagnosed with IRIS Stage 1 CKD when the kidney function most affected is something other than GFR. Examples include primary glomerular disease causing proteinuria, abnormal blood pressure regulation causing hypertension, and signs of a primary tubular disorder (such as glucose in the urine, crystals in the urine, or acidosis). In all of these dogs and cats, serial monitoring of kidney function is warranted to find out whether the disease is progressive and leading to a loss of functioning nephrons.

Primary Glomerular Disease

It is important to understand that severe proteinuria, even to the point of nephrotic syndrome, can be present without any decrease in GFR and therefore without azotemia. Severe glomerular proteinuria is more common in dogs than in cats. It occurs because the glomerular filtration barrier loses its structural integrity, without an initial substantial loss of functioning nephrons.

The presenting signs in patients with primary glomerular disease vary. In some, signs of an underlying disease, such as leishmaniosis, predominate. In others, a protein-losing nephropathy (PLN) is suggested by routine blood and urine tests, which show low blood albumin, high cholesterol, and proteinuria, while the clinical signs (malaise, a poor coat, lethargy) are very nonspecific. Occasionally, patients with PLN are presented with signs of a blood clot (thromboembolism), such as hind-limb weakness or difficulty breathing, or with the consequences of systemic hypertension. A kidney biopsy is indicated in the diagnostic work-up of patients with primary glomerular disease without an identifiable underlying cause, provided they are not azotemic or only mildly azotemic (IRIS CKD Stage 1 or 2).

There is no definitive cut-off for the severity of proteinuria, as measured by the urine protein-to-creatinine ratio (UPC), that distinguishes primary glomerular disease from other kidney diseases, or even from diseases outside the kidney. However, primary glomerular disease is considered likely if the UPC is repeatedly above 2.0 in the absence of urinary tract inflammation, particularly if the patient also has low blood albumin. Normal UPC values are below 0.2 in both dogs and cats, but there is a gray zone above this value. Many of these patients have tubulointerstitial kidney disease that causes glomerular hypertension and secondary glomerular injury. Some have primary glomerular disease, and some with borderline proteinuria (UPC 0.2 to 0.4 in cats, 0.2 to 0.5 in dogs) have an underlying systemic disease or systemic hypertension.

Systemic Hypertension

Although many cats and dogs diagnosed with systemic hypertension have azotemic CKD, some are diagnosed before azotemia develops. When no other cause of hypertension can be identified (such as hyperadrenocorticism, pheochromocytoma, an aldosterone-producing adrenal tumor, or hyperthyroidism), these patients are often presumed to have underlying CKD. Dogs with proteinuria caused by leishmaniosis have been shown to be at increased risk of hypertension even when they are not azotemic (reference 1). Ideally, dogs and cats would be diagnosed with hypertension through routine, pre-emptive blood pressure measurement. Unfortunately, especially in cats, hypertension is often diagnosed only after signs of target-organ damage have developed, particularly eye problems such as bleeding in the eye (hyphema), blindness, or retinal detachment. These seem to be less common in dogs.

Loss of Urine Concentrating Ability

Sometimes a dog or cat is presented for polyuria and polydipsia, and a kidney cause is suspected once other common causes have been ruled out, often because the kidneys look abnormal on ultrasound. Loss of nephron function may interfere with the concentrating gradient in the renal medulla while still not being enough to cause azotemia. This can be an indication to measure GFR. This situation is thought to be more common in dogs than in cats. In cats, owners typically notice increased thirst and urination only once the cat is already azotemic.

Extending this line of thought, it is tempting to assume that a patient with sub-maximally concentrated urine (a urine specific gravity above 1.008 but below 1.030 in dogs, or below 1.035 in cats) and no other obvious cause of polyuria and polydipsia has kidney disease, even if blood creatinine and SDMA are within the laboratory reference range. Some of these patients undoubtedly do have mild kidney disease, but not all. It is important to recognize the differences between species here. Cats typically drink relatively little, and when their kidney function is good they usually produce concentrated urine (above 1.035), especially if they have been fasted or eat only dry food. Dogs, in contrast, drink more and do not always concentrate their urine, and randomly collected samples quite often have a specific gravity between 1.020 and 1.030. Other causes of polyuria and polydipsia, such as hyperadrenocorticism, are also relatively more common in dogs than in cats. Non-concentrated urine in a dog without other clinical signs should therefore not be over-interpreted as a sign of kidney disease. It is also helpful to remember that hyposthenuric urine (specific gravity below 1.008, in a patient that has not been receiving supplementary fluids) does not support a diagnosis of CKD.

Identifying Patients With Mildly Reduced GFR (IRIS CKD Stage 2)

Some dogs and cats with IRIS CKD Stage 2 are easily diagnosed because blood creatinine and SDMA are above the laboratory reference range. However, other patients may also be diagnosed with Stage 2 CKD, including:

• Dogs and cats with independent evidence of CKD, such as structural changes, proteinuria, or glucose in the urine with normal blood glucose, as described above, whose creatinine or SDMA values reach the threshold for Stage 2
• Dogs and cats with an increase in creatinine that is still within the reference range, together with a persistent SDMA concentration of 18 µg/dL or more
• Dogs and cats with a persistently increased SDMA (18 µg/dL or more) but a creatinine within the reference range, especially if the patient is poorly muscled or has some other evidence of CKD from the history, physical examination, or diagnostic imaging

With the last two groups, there is a trade-off between the sensitivity and the specificity of the tests. On one hand, there is a desire to detect CKD as early as possible, because treatment is most likely to help when the number of functioning nephrons is not severely reduced. On the other hand, some patients may be misdiagnosed with CKD. Misdiagnosis is a problem not only because of the unnecessary worry and expense for owners, but also because these patients may be included in trials of new treatments for CKD. To prove that a treatment is useful, there must be a difference in the rate of disease progression between the two treatment groups, or between treatment and placebo. If many patients that are unlikely to have progressive disease are included, the conclusions about the treatment's effectiveness will be wrong. In the future, it is hoped that markers of "active" kidney disease will be developed to complement the indirect markers of GFR (creatinine and SDMA) and help identify the dogs and cats with active disease that treatment should target. This might include mild increases in markers of acute kidney injury.

In some situations, creatinine or SDMA is likely to be mildly increased without CKD being present. Examples are greyhounds and Birman cats, in which GFR appears to be relatively low even in healthy individuals without evidence of kidney disease. In greyhounds, the effect is further confounded by their high muscle mass. These examples can be considered outliers, but reference intervals also differ for other dog breeds, and possibly for cat breeds. Whichever test is used as a surrogate marker of GFR in routine practice, whether blood creatinine, SDMA, or an alternative, its accuracy for diagnosing CKD would probably be maximized by developing breed-specific reference ranges that are narrower than those for the species as a whole.

Even with breed-specific reference ranges, there would still be a great deal of variation in GFR between individuals within a breed. A study of healthy beagles with serial GFR measurements found that, although there was significant day-to-day variation, the variation between individuals was much greater, almost two-fold (reference 8). This means that the best way to show that a patient's GFR has decreased is to use the animal as its own control, by obtaining baseline measurements of creatinine, and ideally SDMA, when it is a young, healthy adult. Preferably, these measurements should be performed in a reference laboratory rather than in-house, and the same laboratory should be used for follow-up testing. However, when individual baseline values are used to monitor a patient, care must be taken not to over-interpret small changes, because day-to-day variation in blood creatinine has been found to be as high as 20% (reference 4).

It is important to recognize that the blood creatinine cut-off between IRIS Stages 1 and 2 is not intended to replace individual laboratory reference ranges. A creatinine (or SDMA) concentration above this threshold does not, on its own, mean that a patient has CKD. A patient must be diagnosed with CKD before the staging system is applied, and markers of GFR alone are sufficient to diagnose CKD only if they are clearly abnormal, or if mild increases are sustained or increasing on serial tests, and the increase is not pre-renal or post-renal in origin.

Although IRIS cut-offs are not meant to replace laboratory reference intervals, it is notable that for many laboratories it is unclear how their reference intervals were derived. Where this is known, reference intervals are often based on relatively small numbers of animals that were not always rigorously confirmed to be healthy. Methods do differ between laboratories, so some differences in reference intervals are to be expected, but it is not clear that the laboratories whose creatinine measurements "run high" are the ones with the highest reference ranges (references 2 and 3). This has also been recognized in human clinical pathology, and it has led to a move toward harmonizing reference ranges between laboratories, using shared standards and consistent adjustments for differences in method. If similar harmonization could be achieved between veterinary laboratories, it would help with serial monitoring of patients and with establishing breed-specific reference ranges (references 5 to 7), because it would allow larger data sets to be combined.

References

1. Cortadellas O, del Palacio MJ, Bayón A, et al. Systemic hypertension in dogs with leishmaniasis: prevalence and clinical consequences. Journal of Veterinary Internal Medicine 2006;20:941–947.
2. Braun JP, Cabé E, Geffré A, et al. Comparison of plasma creatinine values measured by different veterinary practices. Veterinary Record 2008;162:215–216.
3. Ulleberg T, Robben J, Nordahl KM, et al. Plasma creatinine in dogs: intra- and inter-laboratory variation in 10 European veterinary laboratories. Acta Veterinaria Scandinavica 2011;53:25.
4. Jensen AL, Aaes H. Critical differences of clinical chemical parameters in blood from dogs. Research in Veterinary Science 1993;54:10–14.
5. Reynolds BS, Concordet D, Germain CA, et al. Breed dependency of reference intervals for plasma biochemical values in cats. Journal of Veterinary Internal Medicine 2010;24:809–818.
6. Dunlop MM, Sanchez-Vazquez MJ, Freeman KP, et al. Determination of serum biochemistry reference intervals in a large sample of adult greyhounds. Journal of Small Animal Practice 2011;52:4–10.
7. Hall JA, Yerramilli M, Obare E, et al. Comparison of serum concentrations of symmetric dimethylarginine and creatinine as kidney function biomarkers in cats with chronic kidney disease. Journal of Veterinary Internal Medicine 2014;28:1676–1683.
8. Kampa N, et al. Day-to-day variability in glomerular filtration rate in normal dogs by scintigraphic technique. Journal of Veterinary Medicine Series A: Physiology, Pathology, Clinical Medicine 2003;50(1):37–41.