Glomerular Filtration Rate (GFR): What It Means for Your Pet

Adapted from the IRIS education article "Glomerular filtration rate in dogs and cats" by R Heiene (Oslo, Norway) and HP Lefebvre (Toulouse, France), 2013.

Introduction

Glomerular filtration rate (GFR) is the rate at which the kidneys filter blood, and it is considered the single most useful and sensitive test of kidney function (reference 1). Any decrease in GFR generally means that kidney disease is occurring or progressing, so measuring GFR is central to evaluating the severity and progression of kidney diseases, especially chronic kidney disease (CKD). In small animal medicine, the staging system of the International Renal Interest Society (IRIS) is based on blood markers of GFR, creatinine and symmetric dimethylarginine (SDMA), rather than on GFR itself, but the IRIS Board considers that GFR measurement could in future become the main criterion for staging, as it is in people (reference 2). Several methods of measuring GFR have now been validated in dogs and cats, including plasma clearance techniques using non-radioactive markers such as creatinine and iohexol, which are easier to use than urinary clearance methods. These plasma clearance methods make it practical to measure GFR in both clinical and research settings (reference 3). This article reviews developments and current challenges in using plasma clearance methods to assess GFR. For a practical summary of the procedures, see our article on GFR in practice.

Blood Sampling Procedure and Strategy

Plasma clearance methods require repeated blood samples over several hours after intravenous injection of a suitable marker. Clearance (Cl) is then calculated with the formula Cl = GFR = dose / AUC, where AUC is the area under the curve of the marker's plasma concentration plotted against time. After an initial distribution phase, the marker's concentration typically falls exponentially as it is eliminated by the kidneys.

The more blood samples are taken, and the longer the collection period, the more accurate the estimate of the AUC, and therefore of GFR. Ideally, sampling would continue until all of the marker has been excreted, but the test is easier if the number of samples and the sampling period are limited. Much research has focused on finding the best number and timing of samples for practical limited-sampling strategies.

Repeatedly collecting 5 mL blood samples from cats or miniature dogs can be difficult and can cause excessive blood loss. The total volume of blood collected can be reduced if the assay needs only very small volumes of plasma or serum. For example, 0.2 mL of blood is enough for a blood creatinine assay using an enzymatic method. Pediatric capillary collection devices have been proposed as an alternative to vacuum tubes for these tests (reference 4), with potential advantages including no vein collapse, less blood withdrawn, and improved safety.

Blood samples should be timed to minimize the proportion of the AUC that has to be extrapolated (reference 1). The last sample is especially important, because it determines the percentage of the area that is extrapolated, which is calculated from the slope of the elimination phase and the last measured concentration. A slight error in calculating the slope can also cause an error in the clearance. Ideally, the extrapolated area should be less than 20% of the total AUC. The lower the kidney function, the more slowly the marker is eliminated: a slowly excreted marker produces a large AUC, reflecting low kidney function, and a rapidly excreted marker a small AUC, reflecting high kidney function.

Unfortunately, in clinical practice the patient's GFR is unknown, so the best time for the last blood sample is uncertain. For most markers, this is a problem only when GFR is very low, in which case blood creatinine will be increased and late sampling can be planned.

Creatinine has an elimination half-life two to three times longer than most other markers because of its larger volume of distribution. Blood samples at 1 and 10 hours after injection give a relatively accurate estimate of the AUC, with a maximum error of −14% compared with the AUC from an 11-point profile. However, sampling at inappropriate times can lead to errors of more than 400% (reference 6). If the creatinine in the samples can be analyzed immediately and blood creatinine remains high during the day of sampling, keeping the dog in the clinic overnight for a 24-hour sample may be an option.

Correction formulas derived from those used in human medicine, such as the Brøchner-Mortensen formula, have been proposed and applied to dogs and cats to estimate the true GFR from a clearance measured with a limited sampling strategy (references 5 to 10).

Repeated GFR Measurements

A potential problem with repeated GFR measurements in a patient is variability within the individual and between days, that is, the reproducibility of the measurement. If a dog is tested on two occasions, how large must the difference between the two results be to exceed physiological and analytical variability and be considered clinically relevant? Studies in healthy and diseased animals have found between-day coefficients of variation generally below 20% for most markers, which is considered acceptable (references 11 to 14).

Repeated clearance measurements in a patient should use the same technique, to avoid misinterpretation due to differences between methods. Differences of 10% to 20% are often found between the clearances of different markers measured in the same animals at the same laboratory (references 8 and 15 to 18).

Physiological Factors Affecting GFR

Diet and Hydration

Of the dietary factors that can affect kidney function, protein intake is the most important, because a protein-rich meal can increase GFR. Hydration status is also crucial, because dehydration, even when subclinical, reduces GFR. Animals should therefore be well hydrated but fasted overnight before the test.

Age

In growing beagle puppies, mean GFR decreased by about 40%, from 4.1 to 2.5 mL/min/kg, between 9 and 27 weeks of age (reference 19). GFR was 87% higher in 2-month-old beagle puppies than in beagles aged 6 to 9 years (reference 20). These results show that a progressive decline in GFR is normal in growing puppies and should not be interpreted as the development of CKD. In kittens, GFR is higher between 9 and 19 weeks of age than in younger or older animals (reference 21).

In adult dogs, aging seemed to have a limited effect on GFR, or an effect hidden by other influences, in both longitudinal (reference 22) and cross-sectional (reference 5) studies. In cats, the plasma clearance of creatinine, but not of iohexol, was about 25% lower at 9 to 12 years of age than at 7 to 12 months (reference 14). Another feline study found no correlation between age and the plasma clearance of iohexol or creatinine (reference 9). Any decline in kidney function with aging therefore seems limited in healthy dogs and cats, although the clinical relevance of GFR decline in older animals needs further study.

Body Weight

GFR per kilogram is lower in large dogs than in small dogs. In healthy adult dogs, negative correlations have been shown between body weight and GFR assessed by the plasma clearance of creatinine (113 dogs; reference 22) and of iohexol (118 dogs; reference 5).

In cats, too, there was a negative linear relationship between body weight and GFR assessed by the plasma clearance of iohexol and creatinine, although the R² value was low, below 20% (reference 9). In domestic shorthair cats, body weight appears to have a slight effect on blood creatinine (mean ± SD 142 ± 36 µmol/L for cats under 4 kg and 161 ± 22 µmol/L for cats over 5 kg; reference 23).

This effect of size may seem surprising at first, but it is not unexpected, because in several species kidney function has been shown to relate more closely to metabolic needs than to body weight (reference 24).

Breed

Breed effects on blood creatinine have been reported in dogs and cats (references 23 and 24). This does not necessarily mean that GFR depends on breed, because the production of creatinine in muscle also affects blood creatinine. A good example is the greyhound, whose higher blood creatinine has been shown to result from higher creatinine production rather than lower GFR (references 25 and 26).

However, breed differences might still influence GFR. In one study (reference 27), GFR was lower in German shepherds than in English pointers and English setters (2.5 ± 0.7 versus 3.5 ± 0.6 and 3.4 ± 0.8 mL/min/kg, respectively). Age was similar in the three breeds, but the German shepherds were heavier, so differences in body weight, a confounding factor in this study, may have contributed to the differences in GFR.

Indexing GFR

Clearance is measured in mL/min. Standardizing (indexing or scaling) GFR to body size allows the result to be compared with reference ranges for healthy animals, and GFR in dogs and cats is traditionally expressed in mL/min/kg. A current challenge is deciding the most appropriate way to standardize GFR.

Body surface area (BSA) has been proposed as the reference for indexing physiological variables, and it is used routinely in human medicine. However, it has been recommended that indexing GFR to BSA be abandoned (reference 28), and the formulas used to estimate BSA in dogs are probably inaccurate (reference 29). Indexing to extracellular fluid volume (ECFV) has been proposed as an alternative, because one of the kidney's main roles is to regulate the composition of body fluids (reference 30). Nevertheless, standardizing to ECFV did not substantially change the relationship between GFR and body weight in adult dogs (reference 5). Moreover, when GFR was indexed to ECFV, differences between puppies and adult dogs were still seen, but they were reversed (reference 20). In an earlier canine study, standardizing to body weight, BSA, or ECFV produced quite different results and, for some dogs, changed the clinical interpretation of the GFR (reference 10). A similar problem exists in cats: standardizing to BSA gave a larger between-individual coefficient of variation (36%) than standardizing to body weight (27%) or ECFV (24%) (reference 9).

Standardizing GFR is also a concern in obese human patients, who differ clearly from lean people of similar body weight: the heavier the patient, the lower the GFR indexed to BSA. Other ways of indexing GFR have been tested without success, and an absolute, uncorrected GFR is currently recommended in obese patients (reference 31). There are no published data on indexing GFR in obese dogs or cats. In children, variation in body shape and build causes difficulties with prediction formulas and simplified approaches (reference 32).

Further research is needed to find the best method of standardizing GFR in dogs and cats. This research is not simple, because there is no gold standard to compare against. Ideally, regression analysis would be applied to a large, representative population of adult dogs of different breeds, body weights, ages, and sexes to evaluate the different methods of standardization. A simpler alternative would be to divide the canine population into body weight categories and define an average cut-off for GFR in each category. The main limitation is that the wider the weight range in a category, the less accurate its cut-off value. For now, it seems logical to standardize GFR to body weight, because most publications on reference intervals use this approach.

Reference Intervals

Defining GFR reference intervals for dogs and cats is necessary to classify patients as normal or abnormal. Reference values have been published only for adult animals, and values differ in puppies and kittens (see above). The lower cut-off is generally considered to be between 1.5 and 2.0 mL/min/kg. However, given the physiological variability seen in recent studies, a single cut-off does not appear acceptable for the whole canine population. Stratification by body weight, but possibly not by age, is advisable.

In one study of 113 healthy dogs (reference 33), the dogs were divided into four body weight categories: mini, medium, maxi, and giant. Their GFR values (mean ± SD) were 3.7 ± 0.5, 3.0 ± 0.5, 2.5 ± 0.4, and 2.4 ± 0.6 mL/min/kg, respectively (Lefebvre HP et al., unpublished data). If the GFR values were normally distributed, the lower limits of the reference ranges could be considered to be 2.7, 2.0, 1.7, and 1.3 mL/min/kg. In another study of 118 healthy adult dogs, the reference ranges were 1.54 to 4.25, 1.29 to 3.50, 1.2 to 3.36, and 1.12 to 3.39 mL/min/kg for body weight quartiles of 1.8 to 12.4, 13.2 to 25.5, 25.7 to 31.6, and 32.0 to 70.3 kg, respectively (reference 5).

These results, obtained with two different markers (creatinine and iohexol), show that the lower cut-off is higher for small dogs and lower for giant dogs. A single cut-off of 1.5 mL/min/kg for declaring a dog to have impaired kidney function would therefore be too low for small dogs, giving some false-negative results, and too high for giant dogs, giving some false-positive results.

Breed may also affect reference intervals. For example, tentative reference intervals for English pointers, English setters, and German shepherds are 2.3 to 5.1, 1.8 to 5.0, and 1.7 to 3.8 mL/min/kg, respectively (reference 27). It seems impractical to establish breed-specific reference intervals for all but the major breeds. In the authors' experience, most breeds tend to follow their weight category in terms of reference values in healthy dogs.

Plasma clearance reference ranges have been proposed for cats: 1.0 to 3.5 mL/min/kg for iohexol and 1.3 to 3.8 mL/min/kg for creatinine (reference 9). From the point of view of allometric relationships (reference 24), it is interesting that the GFR cut-off is lower in cats than in dogs of similar body weight.

These preliminary results give veterinary nephrologists some estimates of reference intervals, but further studies in larger populations of dogs and cats are clearly needed.

GFR in Cardiovascular and Endocrine Diseases

GFR measurement may also be useful for assessing kidney function in dogs and cats with diseases that do not primarily affect the kidney, as has been shown for some cardiovascular and endocrine diseases.

Cardiovascular Diseases

CKD is prevalent in people with cardiovascular disease, in whom even a very slight decrease in kidney function can dramatically increase cardiovascular risk (reference 34). Azotemia is common in dogs with heart disease, but it is mostly mild to moderate and is not related to changes in cardiovascular variables (reference 35). Interestingly, in 24 dogs with chronic valvular heart disease, GFR was significantly lower in dogs in NYHA classes III and IV (1.7 ± 0.7 mL/min/kg) than in classes I and II (3.1 ± 0.8 mL/min/kg). Only 1 of 15 dogs in classes I and II had a GFR below 2 mL/min/kg, and only 2 of 9 dogs in classes III and IV had a GFR above 2 mL/min/kg (reference 35). Further studies are needed to identify the cause of the decreased GFR and its relevance to the outcome of heart disease.

Endocrine Diseases

Endocrine systems and kidney function interact in both health and disease. One of the best examples in veterinary medicine is probably the effect of hyperthyroidism and hypothyroidism on kidney function in small animals (reference 39).

Hyperthyroidism is a common endocrine disorder in older cats, and pre-existing CKD has been reported in 14% to 40% of affected cats (reference 36). Evaluating kidney function in a hyperthyroid cat is essential, because hyperthyroidism can mask, and sometimes worsen, CKD. CKD became clinically apparent after treatment of hyperthyroidism in about 17% to 39% of cats (reference 36). A challenge in these patients is predicting the risk of azotemia after treatment. Pre-treatment GFR has been shown to predict the development of CKD after treatment (references 37 and 38), and an accurate evaluation of kidney function can be made from 4 weeks after radioactive iodine (I-131) treatment (references 38 and 39).

GFR was decreased in dogs with experimentally induced hypothyroidism, while blood creatinine remained unchanged (reference 40). In 14 dogs with naturally occurring hypothyroidism, GFR was below 2 mL/min/kg in every dog and increased with levothyroxine treatment (reference 41). The possible effects of kidney dysfunction and subclinical CKD on the clinical outcome of hypothyroid dogs need further investigation.

Other situations in which the adequacy of kidney function may be questioned include diabetes mellitus, naturally occurring hyperadrenocorticism, and glucocorticoid treatment. GFR measurement is of limited value at present in these situations, because little is known about how they affect GFR, and studies are needed to establish GFR values for these patients.

References

  1. Heiene R, Lefebvre HP (2007). Assessment of renal function. In: Elliott J, Grauer GF, eds. BSAVA Manual of Canine and Feline Nephrology and Urology, 2nd edition, pp. 117–125.
  2. Elliott J (2007). Staging chronic kidney disease. In: Elliott J, Grauer GF, eds. BSAVA Manual of Canine and Feline Nephrology and Urology, 2nd edition, pp. 159–166.
  3. Lefebvre HP (2010). Glomerular filtration rate in dogs and cats. Proceedings of the ACVIM Forum 2010, pp. 491–493.
  4. Reynolds BS, Boudet KG, Faucher MR, Germain C, Geffré A, Lefebvre HP (2007). Comparison of a new device for blood sampling in cats with a vacuum tube collection system: plasma biochemistry, haematology and practical usage assessment. Journal of Feline Medicine and Surgery 9:382–386.
  5. Bexfield NH, Heiene R, Gerritsen RJ, Risøen U, Eliassen KE, Herrtage ME, Michell AR (2008). Glomerular filtration rate estimated by 3-sample plasma clearance of iohexol in 118 healthy dogs. Journal of Veterinary Internal Medicine 22:66–73.
  6. Brown SA, Finco DR, Boudinot FD, Wright J, Tarver SL, Cooper T (1996). Evaluation of a single injection method, using iohexol, for estimating glomerular filtration rate in cats and dogs. American Journal of Veterinary Research 57:105–110.
  7. Finch NC, Syme HM, Peters AM, Gerritsen RJ, Croubels S, Heiene R (2011). A formula to predict glomerular filtration rate (GFR) using plasma clearance of iohexol in cats. Journal of Veterinary Internal Medicine 2011.
  8. Gleadhill A, Michell AR (1996). Evaluation of iohexol as a marker for the clinical measurement of glomerular filtration rate in dogs. Research in Veterinary Science 60:117–121.
  9. Heiene R, Reynolds BS, Bexfield NH, Larsen S, Gerritsen RJ (2009). Estimation of glomerular filtration rate via 2- and 4-sample plasma clearance of iohexol and creatinine in clinically normal cats. American Journal of Veterinary Research 70:176–185.
  10. Heiene R, Moe L (1999). The relationship between some plasma clearance methods for estimation of glomerular filtration rate in dogs with pyometra. Journal of Veterinary Internal Medicine 13:587–596.
  11. Finco DR, Braselton WE, Cooper TA (2001). Relationship between plasma iohexol clearance and urinary exogenous creatinine clearance in dogs. Journal of Veterinary Internal Medicine 15:368–373.
  12. Finco DR (2005). Measurement of glomerular filtration rate via urinary clearance of inulin and plasma clearance of technetium Tc 99m pentetate and exogenous creatinine in dogs. American Journal of Veterinary Research 66:1046–1055.
  13. Miyamoto K (2001). Evaluation of plasma clearance of inulin in clinically normal and partially nephrectomized cats. American Journal of Veterinary Research 62:1332–1335.
  14. van Hoek I, Vandermeulen E, Duchateau L, Lefebvre HP, Croubels S, Peremans K, Polis I, Daminet S (2007). Comparison and reproducibility of plasma clearance of exogenous creatinine, exo-iohexol, endo-iohexol, and 51Cr-EDTA in young adult and aged healthy cats. Journal of Veterinary Internal Medicine 21:950–958.
  15. Bird NJ, Peters C, Michell AR, Peters AM (2007). Extracellular distribution volumes of hydrophilic solutes used to measure the glomerular filtration rate: comparison between chromium-51-EDTA and iohexol. Physiological Measurement 28:223–234.
  16. Florijn KW, Barendregt JNM, Lentjes EGWM, van Dam W, Prodjosudjadi W, van Saase JLCM (1994). Glomerular filtration rate measurement by "single-shot" injection of inulin. Kidney International 46:252–259.
  17. Frennby B, Sterner G (2002). Contrast media as markers of GFR. European Radiology 12:475–484.
  18. Watson AD, Lefebvre HP, Concordet D, Laroute V, Ferré JP, Braun JP, Conchou F, Toutain PL (2002). Plasma exogenous creatinine clearance test in dogs: comparison with other methods and proposed limited sampling strategy. Journal of Veterinary Internal Medicine 16:22–33.
  19. Lane IF, Shaw DH, Burton SA, Donald AW (2000). Quantitative urinalysis in healthy Beagle puppies from 9 to 27 weeks of age. American Journal of Veterinary Research 61:577–581.
  20. Laroute V, Chetboul V, Roche L, Maurey C, Costes G, Pouchelon JL, De La Farge F, Boussouf M, Lefebvre HP (2005). Quantitative evaluation of renal function in healthy Beagle puppies and mature dogs. Research in Veterinary Science 79:161–167.
  21. Hoskins JD, Turnwald GH, Kearney MT, Gossett KA, Fakier N (1991). Quantitative urinalysis in kittens from 4 to 30 weeks after birth. American Journal of Veterinary Research 52:1295–1299.
  22. Queau Y, Biourge V, Germain C, Braun JP, Watson AD, Jeunesse E, Lefebvre HP (2007). Effect of aging on blood creatinine clearance in dogs. Journal of Veterinary Internal Medicine 21:598 (abstract).
  23. Reynolds BS, Boudet KG, Germain CA, Braun JPD, Lefebvre HP (2008). Determination of reference intervals for plasma biochemical values in clinically normal adult domestic shorthair cats by use of a dry-slide biochemical analyzer. American Journal of Veterinary Research 69:471–477.
  24. Edwards NA (1975). Scaling of renal functions in mammals. Comparative Biochemistry and Physiology A 52:63–66.
  25. Hill RC, Crawford PC, Scott KC, Queau Y, Biourge V, Lefebvre HP (2007). Creatinine disposition and daily production in healthy greyhound dogs. Journal of Veterinary Internal Medicine 21:274.
  26. KuKanich B, Coetzee JF, Gehring R, Hubin M (2007). Comparative disposition of pharmacologic markers for cytochrome P-450 mediated metabolism, glomerular filtration rate, and extracellular and total body fluid volume of Greyhound and Beagle dogs. Journal of Veterinary Pharmacology and Therapeutics 30:314–319.
  27. Seguela J, Queau Y, Murgier P, Mimouni P, Concordet D, Duperron C, Lefebvre HP (2009). Glomerular filtration rate in healthy English Pointer, English Setter and German Shepherd dogs. Journal of Veterinary Internal Medicine 23:764 (abstract).
  28. Turner ST, Reilly SL (1995). Fallacy of indexing renal and systemic hemodynamic measurements for body surface area. American Journal of Physiology: Regulatory, Integrative and Comparative Physiology 268:R978–R988.
  29. Price GS, Frazier DL (1998). Use of body surface area (BSA)-based dosages to calculate chemotherapeutic drug dose in dogs: I. Potential problems with current BSA formulae. Journal of Veterinary Internal Medicine 12:267–271.
  30. Peters AM (1992). Expressing glomerular filtration rate in terms of extracellular fluid volume. Nephrology Dialysis Transplantation 7:205–210.
  31. Delanaye P, Radermecker RP, Rorive M, Depas G, Krzesinski JM (2005). Indexing glomerular filtration rate for body surface area in obese patients is misleading: concept and example. Nephrology Dialysis Transplantation 20:2024–2028.
  32. Hjorth L, Wiebe T, Karpman D (2002). Correct evaluation of renal glomerular filtration rate requires clearance assays. Pediatric Nephrology 17:847–851.
  33. Lefebvre HP, Jeunesse E, Concordet D, Ferré JP, De la Farge F, Laroute V, Giraudel J, Watson AD (2004). Assessment of glomerular filtration rate using plasma exogenous creatinine clearance test: preliminary results in a healthy canine population. Journal of Veterinary Internal Medicine 18:415 (abstract).
  34. Ritz E, McClellan WM (2004). Overview: increased cardiovascular risk in patients with minor renal dysfunction: an emerging issue with far-reaching consequences. Journal of the American Society of Nephrology 15:513–516.
  35. Nicolle AP, Chetboul V, Allerheiligen T, Pouchelon JL, Gouni V, Tessier-Vetzel D, Sampedrano CC, Lefebvre HP (2007). Azotemia and glomerular filtration rate in dogs with chronic valvular disease. Journal of Veterinary Internal Medicine 21:943–949.
  36. van Hoek I, Daminet S (2009). Interactions between thyroid and kidney function in pathological conditions of these organ systems: a review. General and Comparative Endocrinology 160:205–215.
  37. Becker TJ, Graves TK, Kruger JM, Braselton WE, Nachreiner RF (2000). Effects of methimazole on renal function in cats with hyperthyroidism. Journal of the American Animal Hospital Association 36:215–223.
  38. van Hoek I, Lefebvre HP, Peremans K, Meyer E, Croubels S, Vandermeulen E, Kooistra H, Saunders JH, Binst D, Daminet S (2009). Short- and long-term follow-up of glomerular and tubular renal markers of kidney function in hyperthyroid cats after treatment with radioiodine. Domestic Animal Endocrinology 36:45–56.
  39. Boag AK, Neiger R, Slater L, Stevens KB, Haller M, Church DB (2007). Changes in the glomerular filtration rate of 27 cats with hyperthyroidism after treatment with radioactive iodine. Veterinary Record 161:711–715.
  40. Panciera DL, Lefebvre HP (2009). Effect of experimental hypothyroidism on glomerular filtration rate and plasma creatinine concentration in dogs. Journal of Veterinary Internal Medicine 23:1045–1050.
  41. Gommeren K, van Hoek I, Lefebvre HP, Benchekroun G, Smets P, Daminet S (2009). Effect of thyroxine supplementation on glomerular filtration rate in hypothyroid dogs. Journal of Veterinary Internal Medicine 23:844–849.