Detection of early chronic kidney disease in cats

Introduction

Chronic kidney disease (CKD) is estimated to have a prevalence of up to 32% in cats over 12 years of age 1 and has been reported as the second most common cause of death in cats in the United Kingdom aged 5 years and older 2. In people CKD is recognized as a global public health problem, and the importance of intervention strategies focusing on early diagnosis is regarded as being key to tackling this global crisis. However, the major challenge for doctors is making a true diagnosis of early CKD, in particular due to the limitations of serum creatinine as a marker of glomerular filtration rate (GFR). This challenge is shared globally across disciplines; for the veterinarian, early diagnosis of CKD in cats would be a major advantage, as it prompts close monitoring for progression and the timely use of appropriate therapeutic interventions as well as early investigation for, identification and treatment of underlying primary renal disease. It is hoped that the recent availability of novel biomarkers – such as symmetric dimethylarginine (SDMA) – or other approaches using algorithms will help in identifying cats with early kidney disease, and that future research can aid our understanding of the appropriate therapeutic interventions these cats require to slow disease progression. This article briefly summarizes the current research on early diagnosis of feline CKD and how it can be applied in clinical practice.

Pathogenesis and etiology of feline CKD

CKD is simply defined as “the presence of persistent functional or structural abnormalities of one or both kidneys”. Histopathologically, the most common changes are tubulointerstitial inflammation and fibrosis 3. However, the term CKD is non-specific and does not refer to one underlying disease process, but rather to a heterogeneous syndrome which can be defined as a sustained decrease in renal function over at least 3 months.

The widely accepted model of feline CKD development describes an initiation phase in which one or more factors initiates kidney damage, resulting in nephron loss, which leads to self-perpetuating injury to the kidney; this is termed intrinsic progression (Figure 1) 4. Knowledge of these initiating factors can help the veterinarian in identifying appropriate cats to screen for disease. Initiation factors include primary renal disease (including acute kidney injury, or AKI), aging, and environmental factors 4.

Primary renal disease can be categorized as acquired or congenital disease. The most common congenital disease is autosomal dominant polycystic kidney disease, which affects only Persian or Persian cross cats worldwide. Common acquired disease which may be suspected in CKD include renal lymphoma 3, bacterial pyelonephritis, uroliths of the upper urinary tract, chronic viral infection (FIV, FeLV, FIP and feline morbillivirus) 4 and chronic feeding of unbalanced diets 5.

AKI can be defined as a sudden reduction in kidney function resulting in a change in glomerular filtration, urine production and tubular function, and can be initiated by a variety of insults. Although AKI as an initiator of CKD has not been widely studied in cats, in humans it has been shown that an AKI episode increases the risk of subsequently developing CKD, and greater severity of AKI is associated with larger risk 6. In cats, insults to the kidney can be caused by nephrotoxins (e.g., ethylene glycol), neoplasia, infection, sepsis or – perhaps most importantly in the context of CKD – through ischemia. It has been established that tubulointerstitial changes similar to those in feline CKD occur in the later recovery stage of experimentally-induced ischemic AKI in cats 7, providing supportive evidence that AKI, and specifically ischemic AKI, leads to maladaptive repair mechanisms which can lead on to CKD. The possibility that other causes of AKI produce maladaptive repair and then CKD has not yet been explored.

In feline CKD a single primary renal disease is commonly not identified, and it is hypothesized that a combination of factors, including single or repeated AKIs, as well as animal-specific and environmental factors, act cumulatively to initiate CKD 4. Given the increasing prevalence of the disease in older cats 8 research has focused on establishing the link between CKD and aging. Estimates for the prevalence of CKD in cats aged over 12 years range from 32% 1 to 42% 8. The proportion of geriatric cats without CKD gives evidence that CKD is not an inevitability for aging cats, but it is hypothesized that aging compromises the protective mechanisms of the kidneys, making them less likely to recover from a renal insult. It is also speculated that some of the more common diseases of aged cats, such as hyperthyroidism 4, dental disease 9, hypertension 4 and inflammatory bowel disease 10, may adversely affect the kidneys. Finally, it has been suggested that the increased prevalence of CKD over the last few decades could be attributed to changes in the environment including food, vaccination and effects of environmental stress. For example, a recent epidemiological study noted a correlation between the severity of dental disease in cats and the development of azotemia 9. Although it has been established that dietary modification can slow progression of CKD in IRIS stages 2 and 3, evidence for the role of high levels of dietary phosphate as an initiating factor in CKD is lacking. However, recent studies have revealed a possible risk to renal function when healthy adult cats were fed high levels of phosphorus in an inorganic form 11. Further studies are required to better understand the relevance to feline husbandry, but understanding these possible initiating factors can be of use in clinical practice by allowing veterinarians to undertake targeted screening of cats that may be considered to be at greater risk of developing CKD.

Markers of GFR and CKD

GFR is the volume of ultra-filtrate formed in the nephrons of both kidneys per unit of time, and is correlated to the functioning renal mass. Measurement of the plasma clearance of an exogenous marker of filtration, such as iohexol, is the most accurate method of assessing the functional renal mass available to veterinarians. Typically, the estimation of GFR through measurement of a surrogate such as serum creatinine concentration remains the most useful assessment of kidney function in clinical practice.

The primary challenge for veterinarians using serum creatinine to diagnose early kidney disease in cats is the curvilinear relationship between serum creatinine concentration and GFR, as shown in Figure 2. A substantial decrease in GFR is required before a significant increase in serum creatinine concentration (and consequently azotemia) is found on clinical biochemistry, which is why it is an insensitive indicator of GFR.

Increases in serum creatinine in early disease are small and often still within the laboratory reference interval. The IRIS CKD Staging system defines stage 1 CKD as a non-azotemia animal (serum creatinine concentration < 140 µmol/L, 1.6 mg/dL in cats) with some other renal abnormality present: i.e., persistent inadequate urine concentrating ability with no identifiable non-renal cause, abnormal renal palpation or imaging, persistent proteinuria of renal origin, abnormal renal biopsy results, or an increase in blood creatinine concentration in serially collected samples ¹. However, the identification of cats with IRIS stage 1 and stage 2 (where creatinine is 140-250 µmol/L, 1.6-2.8 mg/dL), when creatinine may be within the laboratory reference interval (and documentation of the creatinine trend is unremarkable), and where evaluation for other clinical evidence of CKD is required, can be challenging.

¹ www.iris-kidney.com/pdf/IRIS_CAT_Treatment Recommendations_2019

This challenge is exacerbated by non-renal factors such as muscle mass 12, age and breed (e.g., the Birman) 13 have all been shown to influence creatinine levels. To allow for these limitations, it is recommended that creatinine is always measured on a fasted blood sample and interpreted considering the breed, muscle mass and age of the cat.

To tackle the limitations of serum creatinine concentration as a marker of early kidney disease, research in recent years has focused on other novel biomarkers of decreased GFR, tubular and glomerular damage that may detect changes earlier in the disease. The most readily available of these to veterinarians is symmetric dimethylargnine (SDMA).

SDMA – what do we know?

Symmetric dimethylarginine (SDMA) is a methylated form of arginine which is found in all intracellular proteins and released into the circulation during protein catabolism. Ninety percent is excreted by the kidneys, and SDMA has been shown to be a surrogate marker of GFR 14. Since 2015 a commercial quantification of SDMA has been available in many countries, whereby serum or plasma SDMA concentration is quantified through a patented immunoassay which has been shown to have good agreement with the gold standard liquid chromatography mass spectrometry (LC-MS) methodology 15.

Serum SDMA concentration has been reported to detect a decrease in GFR before serum creatinine concentration is elevated (as based on the established reference limits), and is now recognized as a useful screening tool for detection of early CKD. In a study of 21 geriatric colony cats with naturally occurring CKD, serum SDMA concentration was elevated above 14 µg/dL an average of 17 months prior to elevation of creatinine above the upper reference interval of 186 µmol/L (2.1 mg/dL) in 17 of the 21 cats 16.

Furthermore, SDMA is also reported to be a highly specific biomarker of reduced GFR, potentially having fewer non-renal influences than creatinine. Although it may be expected that there will be some small biological and individual day-to-day variability, there is evidence that SDMA is not significantly influenced by muscle mass 16 17 or recent protein ingestion 17. Age and breed have been shown to have some effect on SDMA concentrations, with research ongoing to establish age and breed specific reference intervals. Currently it is known that an SDMA concentration of up to 16 µg/dL may reflect normal renal function in juvenile cats 18 and it has been reported that SDMA concentrations are increased in Birman Cats; a breed-specific reference interval for these cats of 3.5-18.7 µg/dL has been suggested.

Given that SDMA is a relatively novel biomarker, understanding of possible non-renal influences on circulating concentrations is still developing; specifically it is important for the veterinarian to consider the influence of drug administration and concurrent disease. It has been reported that the presence or absence of myxomatous mitral valve disease (MVD) and the signs of (or pharmacological treatment for) congestive heart failure have no association with serum SDMA concentrations in dogs 19. Although MVD is specific to dogs, in cats the presence of hypertrophic cardiomyopathy has also been reported to have no effect on SDMA concentration 20, providing preliminary evidence that cardiac disease does not affect SDMA concentrations across species. One study in dogs has indicated that a large tumor burden without a reduction in renal function could result in elevated SDMA 21, and until further studies have been undertaken it should be assumed that this could be the case in cats as well. There is preliminary evidence that feline nephrolithiasis may increase SDMA above the reference interval, although this may be attributed to early alteration in renal function rather than a non-renal influence. Conversely, significantly lower SDMA concentrations have been reported in cats with diabetes mellitus undergoing insulin therapy 20 and in untreated hyperthyroid cats 22. Such findings should be kept in mind when evaluating renal function in cats with these endocrinopathies. In the study of hyperthyroid cats, SDMA had poor sensitivity (33.3%) for predicting development of azotemia following treatment for hyperthyroidism, although it was highly specific (97.7%). This suggests that an elevated SDMA prior to treatment for hyperthyroidism is a good indicator of post-treatment azotemia, but that normal SDMA does not rule out disease.

Markers of glomerular and tubular damage

Whereas serum creatinine concentration and SDMA are surrogate markers of renal function (i.e., GFR), glomerular or tubular damage or dysfunction can be indicated by urinary markers. Several have been identified in veterinary medicine.

Proteinuria is a commonly used marker of glomerular or tubular damage or dysfunction. This is routinely identified in practice by the dipstick colorimetric test which detects urinary albumin (Figure 3), however, it should be noted that both false negatives and, in particular, false positives are common in cats. Having detected proteinuria on a dipstick, pre- and post-renal causes such as hemoglobinuria or a urinary tract infection should be ruled out, and proteinuria should be quantified using the gold standard urine protein:creatinine ratio (UPC). Once persistent proteinuria is confirmed it should be staged according to IRIS guidelines. Even a low magnitude of proteinuria is associated with the development of azotemia, highlighting the importance of including urinalysis when screening cats for early CKD (Figure 4).

Proteinuria may result either from the normal renal handling system of protein being overwhelmed (increased protein loss across the glomerulus) or malfunctioning (reduced ability of the tubular cells to reabsorb filtered proteins). In the healthy kidney, low molecular weight (MW) proteins (< 40 KDa) are able to pass freely through the glomerular filtration barrier; those of intermediate MW (40-69 kDa) have variable permeability depending on their charge, whilst high MW proteins (> 70 kDa) are generally restricted due to their size. Healthy proximal tubule cells reabsorb proteins that are filtered into the tubular space via receptor-mediated endocytosis. If the glomerulus is damaged, permeability of the filtration barrier is increased, resulting in marked proteinuria. Tubular damage will also result in proteinuria from a combination of leakage of proteins from damaged tubular cells, decreased reabsorption of proteins, and upregulation of proteins involved in injury and repair. Apart from albuminuria, other intermediate or low MW proteins may be developed as markers of early CKD in the future. Transferrin, which has a similar molecular weight to albumin but with a different isoelectric point, is reported to be found at very low concentrations in the urine of normal cats, but is increased in the urine of healthy or stage 1 CKD cats in which subsequent renal biopsy confirmed a chronic interstitial nephritis, suggesting it could be a very specific marker of early renal damage 23. Research into low molecular weight proteins such as retinol binding protein and neutrophil gelatinase-associated lipocalin is ongoing.

Urinary proteomics has the potential to identify low molecular weight proteins which may facilitate early diagnosis of CKD in the cat 24. Prospective longitudinal studies to identify and validate urinary markers of early feline CKD are required before these tests are likely to be available in clinical practice.

Diagnosis of CKD and machine learning

Machine-learning models, whereby algorithms are used to analyze data, have been developed in human medicine to assess patient risk, predict individual outcomes and recommend personalized treatments, and it is likely that similar uses will be applied in veterinary medicine in the future. Machine learning has recently been used to develop an algorithm that combines age, urine specific gravity, serum creatinine and urea, collected on at least three occasions from cats undergoing routine health screening, to predict the risk of developing azotemic CKD within one year 25. Interestingly, this study reported that the algorithm performed with a specificity of over 99% and a sensitivity of 63% for predicting cats at risk of CKD a year before the condition was diagnosed by more conventional methods.

Practical diagnosis of early CKD

Clinical presentation

Cats in the later stages of CKD – i.e., more advanced IRIS stage 2 and in IRIS stage 3 and 4 – will often present with polyuria and polydipsia as well as non-specific clinical signs, including weight loss, decreased appetite and lethargy. Physical examination findings may include small kidneys on palpation, which may be irregular in outline, or there may be one enlarged and one small kidney, for example in cases of renal lymphoma or acute ureteral obstruction with resulting hydronephrosis. Cats in early CKD may not present with clinical signs of disease and physical examination findings can also be within normal limits (Figure 5); mild azotemia, an elevation in SDMA or proteinuria may be noted on routine pre-anesthetic screening or as part of a diagnostic work-up for a concurrent condition. Diagnostic testing for CKD, including a biochemistry profile, hematology and urinalysis, can be performed at wellness and vaccination visits of geriatric cats, as well as those cats where the veterinarian feels that the risk of CKD is increased through exposure to the initiating factors discussed above.

Diagnostic testing

Early diagnosis of CKD in cats requires the combined use of plasma or serum creatinine, SDMA concentrations and urinalysis, rather than considering one parameter in isolation, as no single test is 100% specific and sensitive. Upward trends in creatinine, elevation in SDMA above the reference interval, a decline in USG and identification of proteinuria can all be used to aid diagnosis, and should be interpreted according to IRIS guidelines. Renal imaging should also be undertaken following abnormalities noted on palpation or where identified on blood and urine testing. The two following clinical scenarios give practical examples of early CKD diagnosis.

Case study 1

Signalment

“Minnie” Domestic Shorthair, female neutered, 13 years of age.

History

Over last 6 months owner noted worsening polyphagia, weight loss and generally poor coat condition.

Clinical signs

Abnormal physical exam findings included tachycardia, a body condition score (BCS) of 3/9 (Figure 6), weight loss (500 g in 6 months) and an anxious demeanor. Blood pressure via Doppler was 124 mmHg.

Initial diagnostics   

Significant biochemistry tests results (normal values are shown in (Box 1)) included thyroxine (T4) 150 nmol/L; creatinine 106 µmol/L; urea 7 mmol/L; SDMA 17 µg/dL. Urinalysis was unremarkable, but USG was 1.027.

Treatment

Treatment for hyperthyroidism was initiated with thiamazole 2.5 mg q12H PO. After 4 weeks on treatment, Minnie was no longer polyphagic. On clinical exam, the tachycardia had resolved and she had gained 250 g, with a BCS 5/9 (Figure 7). Blood test results were; T4 36 nmol/L; creatinine 120 µmol/L; urea 8.4 mmol/L; SDMA 17 µg/dL. Urinalysis was unremarkable, but USG was 1.025.

Follow-up diagnostics

To follow up on the elevated SDMA noted on the second test once the hyperthyroidism was controlled, a further blood test to check renal parameters was taken two weeks later and revealed creatinine 122 µmol/L; urea 8.8 mmol/L and SDMA 18 µg/dL. Urinalysis was unremarkable, but USG remained low at 1.025. A diagnosis of CKD stage 1 was made due to the persistent elevation in SDMA; this was supported by a USG persistently below 1.035.

8 weeks after confirming IRIS stage 1 CKD the renal parameters were checked again to monitor for progression of CKD and revealed creatinine 204 µmol/L; urea 6.8 mmol/L and SDMA 18 µg/dL. Urinalysis was unremarkable, but USG was 1.019.

Case discussion

Minnie presented with the clinical signs of hyperthyroidism and diagnosis was confirmed by measuring total thyroxine levels. Prior to treatment for hyperthyroidism her creatinine was within normal limits and her urinalysis was unremarkable. However, her SDMA was mildly elevated and her USG was below 1.035, giving the veterinarian an indication that early CKD may be a possibility. However, regardless of pre-treatment values, renal parameters should always be monitored closely alongside treatment for hyperthyroidism, and routine blood and urine tests were repeated 4 weeks after commencing thiamazole treatment. These confirmed that Minnie’s hyperthyroidism was controlled and that whilst the serum creatinine remained within the reference range, the SDMA continued to be elevated.

To confirm persistent elevation of SDMA with controlled hyperthyroidism, renal biochemistry was repeated 2 weeks later. With SDMA elevated on two consecutive occasions two weeks apart Minnie was diagnosed with early CKD and staged at IRIS stage 1; this diagnosis was supported by a USG persistently below 1.035. It was advised that further investigations, including a repeat urinalysis and renal imaging to check for underlying renal disease, should be carried out.

Staging Minnie at stage 1 CKD prompted the veterinarian to monitor her closely for progression of CKD, and 8 weeks after the initial CKD diagnosis her renal parameters revealed azotemia with a USG of 1.019. Stage 2 CKD was diagnosed, and appropriate management according to IRIS guidelines was commenced.

Case study 2

Signalment

“Jeremy”; Norwegian Forest cat, male neutered, 12 years of age (Figure 8)

History

Been in owner’s possession since a kitten, fully vaccinated, seen for routine booster vaccination. Owner has no concerns.

Clinical signs

Physical examination normal.

Systolic BP via Doppler 130 mmHg.

Initial diagnostics

Annual hematology and biochemistry tests (including T4) were performed according to guidelines for geriatric cats. Biochemistry revealed creatinine 135 µmol/L; urea 8 mmol/L; SDMA 18 µg/dL (normal values as shown in (Box 1)) whilst hematology was unremarkable. Urinalysis was also unremarkable, with USG 1.040. With SDMA elevated above the RI a repeat biochemistry was indicated.

Follow-up diagnostics

Jeremy was seen again 4 weeks later to check his renal parameters, although urinalysis was not obtained at this visit. Biochemistry revealed creatinine 130 µmol/L; urea 8.7 mmol/L; SDMA 13 µg/dL. SDMA was not elevated on this occasion and no further action was required.

Case discussion

Elevated SDMA on one occasion in a non-azotemic cat should not be considered diagnostic; SDMA must be persistently elevated on a follow-up testing to allow a diagnosis of early CKD. In Jeremy’s case the USG of 1.040 was also less indicative of early CKD. USG below 1.035 is sometimes taken to indicate reduced urine concentrating ability, but on a spot sample such a finding has poor specificity for renal dysfunction unless combined with other indicators. However, a USG that is > 1.035 on a spot sample makes a diagnosis of early stage CKD less likely, as it indicates urine concentrating ability is adequate. For Jeremy, no further action was required after the follow-up test, and annual monitoring should continue at the next vaccination appointment.

What to do following diagnosis – interventions in early CKD

IRIS guidelines outline the introduction of a renal diet in stage 2 CKD² and feeding a phosphate and protein-restricted diet to cats with azotemic CKD has been shown to improve survival and slow progression of disease 26. There is currently less research into the potential benefits of a similar diet in early or stage 1 CKD. Dietary intervention studies in geriatric cats with IRIS stage 1 CKD fed a test diet containing functional lipids, antioxidants and high-quality protein have demonstrated that such a diet resulted in significant decreases in variable combinations of renal function markers, including SDMA and creatinine, when compared to the normal diets (i.e., those chosen by owners) 27. The study authors speculate that improvement in renal function secondary to the effect of the test diet may explain the stability of, or decrease in, circulating SDMA concentrations. However, no clearance technique was performed to confirm this or to evaluate the significance of changes in serum creatinine concentration in the face of stable serum SDMA or vice versa. It is also worth noting that although creatinine and SDMA can aid in early diagnosis of CKD, both are surrogate markers of GFR only and do not inform on the metabolic status of an animal.

² www.iris-kidney.com/pdf/IRIS_CAT_Treatment_Recommendations_2019

Chronic kidney disease-mineral and bone disorder (CKD-MBD), resulting in derangements of parathyroid hormone (PTH), fibroblast growth factor-23 (FGF23), 25-dihydroxyvitamin D, serum calcium and phosphate, with accompanying renal osteodystrophy and vascular/soft tissue calcification is recognized in cats. Neither creatinine nor SDMA alone are able to inform about the presence of CKD-MBD, and further research is therefore required to establish the derangements in phosphate homeostasis that can be identified with stage 1 CKD and the role of measuring markers of bone mineral disturbance (such as FGF23), in determining which cats require clinical intervention in the form of dietary modification. Currently, FGF23 assessment is not commercially available.

Chronic kidney disease has a significant prevalence in the feline population and is a major cause of death in older cats. Early diagnosis of CKD is clearly advantageous, as it prompts close monitoring for progression and the timely use of appropriate therapeutic interventions. Using serum creatinine concentration remains the most common assessment of kidney function in practice, but the recently developed SDMA test may detect early signs of CKD some months prior to elevation of creatinine above the upper reference interval. However, accurate early diagnosis requires the combined use of plasma or serum creatinine, SDMA concentrations and urinalysis, rather than considering one parameter in isolation, as no single test is 100% specific and sensitive.