Asubstantial number of elderly cats develop renal disease. Survival rate is variable, and although some cats die soon after diagnosis, others have a prolonged survival time.1–3 The estimation of GFR may allow the early identification of renal disease, thereby allowing earlier institution of renoprotective measures such as dietary or medical treatments.4,5 Indications for the measurement of GFR include screening for the presence of renal disease in animals with nonazotemic polyuria or mild increases in plasma creatinine concentration, presurgical or posttreatment monitoring of patients in various clinical situations, screening in breeds with a predisposition to familial renal disease, and dosage guidance for administration of renally excreted drugs.6 Sequential measurements of GFR may also help to evaluate of the effects of therapeutic interventions on renal function over time.
Glomerular filtration rate is regarded as the best overall index of renal function in health and disease and is optimally estimated by measurement of clearance of a marker substance.7,8 The urinary clearance of the fructose polymer inulin has long been regarded as the reference method for determining GFR in humans, dogs, and cats.
As an alternative to urine collection procedures in cats, determinations of the plasma clearance of a marker (eg, inulin), the iodine-containing radiographic contrast medium iohexol, radionuclides, and exogenous creatinine have been evaluated.9–21 The expense, lack of availability of inulin, requirement for transport of frozen plasma samples, and cumbersome laboratory analyses generally preclude the use of inulin. The use of radionuclides requires access to a nuclear medicine facility. The radiographic contrast medium iohexol has been extensively used in nephrologic evaluations in humans.22,23 However, contrary to findings in humans in which creatinine undergoes more extensive interaction with various body systems,24 plasma clearance of exogenous creatinine seems to produce reliable GFR estimates in dogs and cats.14,25 Currently, the most practical markers for GFR assessment appear to be creatinine and iohexol. The advantage of creatinine as a marker is that analysis of plasma creatinine concentration is widely available. The disadvantages are that the elimination time of creatinine from the body is 3 times longer than times for most other markers. Furthermore, there is a lack of commercially available injectable forms of creatinine, although it can be produced in a pharmacy. The advantages of iohexol include its stability, which allows samples to be sent via mail to an external laboratory for analysis, and its generally low toxicity and cost. Iohexol is also widely available because it is commonly used as a radiographic contrast agent, and it is gaining popularity in human medicine for estimation of GFR. The disadvantage is that laboratory analysis of plasma concentrations of iohexol is not widely available.
Because of their typically small body size and potential problems in handling, it can be difficult to obtain multiple blood samples from cats. Interpretation of previous method validation studies for GFR estimation in cats is sometimes difficult because of the small number of cats evaluated or because of the small number of blood samples collected from each animal. In 1 study,10 after plasma was harvested from blood samples collected from each of 5 cats, the remaining erythrocytes were washed and reinjected into the donor. In other studies,10,13,19 sample collection for plasma clearance determination was terminated before an optimal proportion of the area under the plasma disappearance curve was defined.
Limited sample collection strategies for plasma clearance procedures are used in human and veterinary medicine for a simple and clinically acceptable estimation of GFR. However, correction formulae must be applied when these strategies are used; that which is most widely used is the Brøchner-Mortensen formula.26 To our knowledge, there is no cat-specific formula to correct for the initial distribution phase, but several investigators have used the human correction formula in their studies in cats.10,13,17,19
Morphologic changes in human and canine glomeruli are associated with aging,27,28 and in humans, there is an associated decline in GFR. Limited data exist to indicate a corresponding decline in GFR in aging cats. Although not extensively studied, some published data indicate differences in GFR among mammalian species of different body sizes.29 A dependency of GFR on body weight was also detected in dogs of various breeds and sizes.30
The purpose of the study reported here was to compare 2 methods for estimating GFR in cats, evaluate the effect of ages and body size and their interaction on GFR estimates, and establish reference ranges for GFR (derived via the different methods) in clinically normal cats. The accuracy of GFR estimates derived when the number of blood samples evaluated was decreased from 4 to 2 was also evaluated.
Materials and Methods
Cats—The study was conducted at Specialistencentrum DeKompaan, Ommen, The Netherlands, in 2006. Ethical approval was in accordance with national regulations, and informed written consent was obtained from the owners. The study population consisted of clinically normal client-owned cats. Mean age of the cats was 5.7 years (range, 1 to 17 years), and mean weight was 4.5 kg (range, 2.4 to 7.5 kg). British shorthair (n = 11), Ragdoll (11), domestic shorthair (10), Devon Rex (8), and Sphynx (8) were the most common breeds; other breeds included Birman (4), Abyssinian (1), and Siamese (1). All cats were considered clinically normal by the owners at the time of the investigation, and this was supported by findings of physical examination and clinicopathologic analyses (20 variables assessed). Exclusion criteria were a previous history of renal or urinary tract disease or abnormal clinicopathologic findings. Two cats with serum creatinine concentrations that were > 20% higher than the upper reference limit were excluded from the calculation of reference ranges, as were the 4 Birmans, of which 2 were azotemic. None of the cats that contributed data for reference range calculation had serum creatinine concentration > 5% higher than the upper limit of the laboratory reference range.a However, all 57 cats were included in the data set used for method comparisons.
Clearance procedure—For each cat, food was withheld for 12 hours prior to the procedure, but free access to drinking water was allowed throughout the study. Iohexolb (647 mg/kg [300 mg of iodine/kg]) was administered IV via a catheter placed in a cephalic vein. The dose of creatinine was 40 mg/kg and was prepared from a solution that had a concentration of 80 mg/mL. The syringes were weighed before and after administration of creatinine to maximize the accuracy of the calculations. Iohexol and creatinine were administered separately as boluses (duration of injection, approx 30 seconds), and each injection was followed by administration of a flush of heparinized saline (0.9% NaCl) solution. The dose of iohexol was measured from the syringe volume, whereas the prepared creatinine syringes were weighed before and after aspirating the injection volume. The completion of the injections represented time zero. Five milliliters of blood was collected from a jugular vein into a tube containing lithium heparin before and at approximately 2, 3, 4, and 5 hours after injection. The exact time of sample collection was used in the calculations. Blood samples were centrifuged for 10 minutes; plasma was collected and frozen at −20°C until analysis.
Laboratory analysis of plasma iohexol concentrations—Plasma concentrations of iohexol stereoisomers exo-iohexol and endo-iohexol were determined via a high-pressure liquid chromatography method with UV detection that has been previously used in cats.31
Laboratory analysis of plasma creatinine concentrations—All assays were performed on the same day. Plasma specimens were thawed for 30 minutes at room temperature (approx 20°C), and creatinine assays were performed by use of an enzymatic method in a dry-slide–technology analyzer.c Quality control was based on repeated measurements of control solutions.d According to the manufacturer, their values can be traced to standard reference materials available from the National Institute of Standards and Technology. Repeatability and within-laboratory imprecision of the creatinine analysis were estimated from assessment of the same batch of control solutions in 10 consecutive replicates and via weekly single measurement over a period of 10 consecutive weeks, respectively. Repeatability and within-laboratory imprecision coefficients of variation were each < 3.1%.
Pharmacokinetic analysis—Clearance (CL1comp) was calculated by use of an equation as follows:
where D = dosage. A commercially available pharmacokinetics computer programe was used to calculate the AUC with a 1-compartmental model; in that model, AUC was calculated as A divided by α, where A is the 0-intercept and A is the rate constant of the terminal monoexponential slope of the curve, as defined by samples collected at 2, 3, 4, and 5 hours.
A calibration procedure was performed to predict the complete-curve plasma clearance, which was based on 10 samples collected over a 6-hour period (CLtrap). The procedure required the use of the empirical polynomial regression formula derived in dogs32 as follows:
The clearance value then was standardized to body weight in kilograms.
Similarly, a calibration procedure was performed to predict the complete-curve GFR/ECFV (ie, total plasma clearance of iohexol standardized to liters of ECFV based on 10 samples collected over a 6-hour period [GFR/ ECFV2comp]) by use of the terminal elimination rate constant k10 defined by samples collected at 2, 3, and 4 hours (GFR/ECFV1comp)32; the procedure required the use of the empirical regression formula32 as follows:
Iohexol standardization to BSA (m2) was also performed by use of an equation as follows:
Glomerular filtration rate estimates predicted by use of the dog-specific correction formula from the 4-sample values also were compared with the estimates predicted by use of the Brøchner-Mortensen formula derived from humans (Clearance = 0.990778 × Cl1 − 0.001218 × Cl12, where Cl1 is the 1-compartmental model clearance value, CL1comp).26
Also, the same calculations were performed with the AUC derived from the line defined by 2 samplesf rather than 4 samples; for iohexol, the 2- and 3-hour samples were used, and for creatinine, the 3- and 5- hour samples were used. Also, the effects of the alternate sampling times were tested by use of the 2- and 3- hour samples for the calculation of 2-sample creatinine clearance and by use of the 3- and 5-hour samples for calculation of 2-sample iohexol clearance.
Statistical analysis—All results are expressed as mean values or 95% confidence intervals constructed by the Student procedure.33 As an index of dispersion, the SD values are provided.
The agreement between CLtrap and CL1comp for iohexol and for creatinine as well as agreement between methods involving 2 or 4 samples for both markers was evaluated via agreement analysis.34,35 This was done in a stepwise manner by first testing the difference between 2 methods against 0 by use of a paired Student t test.36 Then the regression line was tested for deviation from the line of equality. The mean of the mean value between the 2 methods was designated as MeanAve, and the SD of the differences was designated as SDDiff. The limits of agreement were defined as the difference between methods ± 2SDDiff. The standardized agreement index was calculated as 1 – (2SDDiff / MeanAve).37,38 The agreement index was classified according to consensus.39 Outliers were defined as differences outside the limits of agreement. Finally, the correlation coefficient between the mean of the 2 methods and the absolute value of the differences was calculated and tested to evaluate whether the difference changed with increasing clearance values. Reference ranges are expressed as mean ± 2SD.
Simple linear regression and Pearson moment correlation33 were used to study relationships between variables. The coefficient of variation (SD/mean • 100) was calculated as a percentage for each standardization method. All tests in the analyses were performed in a 2-tailed manner with significance set at a value of P ≤ 0.05.
Results
No adverse clinical signs were observed during or after the administration of iohexol in any cat. At 5 hours after injections, 6% and 38% of the total AUC was determined by extrapolation for iohexol and creatinine, respectively.
The values of CLtrap and CL1comp were significantly (P < 0.01) higher for creatinine than for iohexol, with the uncorrected results for the 2 markers differing most (Table 1). The regression between the clearances of creatinine and iohexol differed significantly from the line of equality for both CLtrap and CL1comp. This was also most pronounced for the CL1comp values, for which the correlation between the absolute difference and the mean of the 2 methods was significant (P < 0.01). Regardless of the difference in measurement level, the agreement index between CLtrap of creatinine and CLtrap of iohexol was classified as very good with an index of 0.71 (Figure 1).
Comparison of plasma clearance of iohexol (CLiohexol) and creatinine (CLcreatinine) determined by use of uncorrected CL1comp and the corrected (predicted) CLtrap and by use of predicted clearance of each marker via 4- or 2-sample assessment methods in 51 clinically normal cats.*
| Variable | lohexol and creatinine clearance | Predicted clearance by use of 2 samples vs 4 samples | ||
|---|---|---|---|---|
| CL1comp | CLtrap | CLtrap for iohexol | CLtrap for creatinine | |
| MeanAve (mL/min/kg) | 2.81 | 2.31 | 2.16 | 2.41 |
| SDDiff (mL/min/kg) | 1.06 | 0.67 | 0.68 | 0.68 |
| MeanDiff of CLcreatinine − CLiohexol | 0.46 | 0.28 | 0.02 | 0.07 |
| or CL4S − CL2S (mL/min/kg) | ||||
| Limits of agreement (mL/min/kg) | −0.68 to 1.60 | −0.40 to 0.96 | −0.18 to 0.22 | −0.38 to 0.54 |
| Percentage of outliers | 6 | 6 | 6 | 4 |
| Agreement index | 0.59 | 0.71 | 0.91 | 0.81 |
| Correlation between absolute differences and MeanAve of the 2 methods | 0.45 | 0.15 | 0.10 | 0.19 |
| P value for the correlation between the absolute difference and MeanAve | < 0.001 | 0.24 | 0.44 | 0.14 |
Cats were injected IV with 647 mg of iohexol/kg (300 mg of iodine/kg) and 40 mg of creatinine at 0 hours; samples were collected before and at 2, 3, 4, and 5 hours after injection. For the 2-sample assessment of iohexol clearance, the 2- and 3-hour samples were used; for the 2-sample assessment of creatinine clearance, the 3- and 5-hour samples were used. A value of P ≤ 0.05 was considered significant.
* Although 57 cats were included in the study, 6 cats that were azotemic, Birman cats, or both were excluded from these analyses.
MeanAve = Mean of the mean values derived by use of the 2 methods. SDDiff = SD of the differences between the 2 methods. MeanDiff = Mean difference between the 2 methods. CL4S = Plasma clearance by use of the 4-sample method. CL2S = Plasma clearance by use of the 2-sample method.
Bland-Altman plots of the difference in plasma clearance of creatinine (Cr) versus plasma clearance of iohexol (Ix) against the means of clearances determined by use of a noncompartmental corrected (predicted) trapezoidal model (A) and by use of a 1-compartment model (B) and the difference in plasma clearance of iohexol (C) or creatinine (D) determined by use of a 4-sample method (4S) and by use of a 2-sample method (2S) in 57 cats. Cat were injected IV with 647 ng of iohexol/kg (300 mg of iodine/kg) and 40 mg of creatinine at 0 hours; samples were collected before and at 2, 3, 4, and 5 hours after injection. For the 2-sample assessment of iohexol clearance, the 2- and 3-hour samples were used; for the 2-sample assessment of creatinine clearance, the 3- and 5-hour samples were used. The horizontal line (Mean) represents the mean difference between the 2 methods; dashed lines represent the upper and lower 2 SD limits.
Citation: American Journal of Veterinary Research 70, 2; 10.2460/ajvr.70.2.176
In the 51 nonazotemic cats, mean CLtrap for iohexol was 2.26 mL/min/kg and mean CLtrap for creatinine was 2.55 mL/min/kg, as corrected by the dogspecific formula (Table 2). The corresponding reference ranges were 1.02 to 3.50 mL/min/kg for iohexol and 1.27 to 3.83 mL/min/kg for creatinine. In all 57 cats, the mean clearance for exo-iohexol and endoiohexol was 2.2 mL/min/kg and 1.5 mL/min/kg, respectively.
Mean ± SD and reference ranges (values of mean ± 2 SD) for GFR estimates based on plasma clearance of iohexol and creatinine determined by use of an uncorrected CL1 comp and corrected (predicted) CLtrap in 51 clinically normal cats (1 to 17 years old) and the 95% confidence Intervals for the lower and upper limits of the reference range.
| Method | Mean ± SD (mL/min/kg) | Reference range for clearance (mL/min/kg) | 95% confidence interval for the lower reference limit (mL/min/kg) | 95% confidence interval for the higher reference limit (mL/min/kg) |
|---|---|---|---|---|
| Iohexol CL1comp | 2.71 ± 0.97 | 0.77–4.65 | ||
| Iohexol CLtrap | 2.26 ± 0.62 | 1.02–3.50 | 0.85–1.26 | 3.26–4.84 |
| Creatinine CL1comp | 319 ± 111 | 0 97–5 41 | ||
| Creatinine CLtrap | 2.55 ± 0.64 | 1.27–3.83 | 1.06–1.58 | 3.20–4.76 |
No significant differences in mean calculated GFR values were detected between the 2- and 4-sample methods for iohexol or creatinine (Figure 1; Table 1). The regression between the 2- and 4-sample clearance values did not differ significantly from the line of equality for either of the 2 methods. Additionally the agreement index for both methods was > 0.80 (classified as very good). The correlation between the absolute difference and the mean of the 2- and 4-sample values was close to zero for iohexol. For creatinine, this correlation was slightly positive (P = 0.09).
No correlation was detected between the CLtrap values for iohexol or creatinine and age (Figure 2). The CLtrap of iohexol and of creatinine both decreased significantly (P ≤ 0.02) with increasing body weight, which explained 12.0% (R2 = 0.12) and 17.9% (R2 = 0.179) of the variation, respectively (Figure 3). Standardization of GFR to ECFV (rather than to weight) did not result in substantial changes in the relationships between GFR estimates and age or weight. Overall, there was no significant correlation between age and weight in this data set. The relationships between the CLtrap values of iohexol or creatinine and plasma creatinine concentration were also assessed (Figure 4). When reference iohexol clearance was standardized to weight, ECFV, or BSA, the coefficient of variation was 27%, 24%, or 36%, respectively.
Relationship between age and estimated GFR determined on the basis of plasma clearance of iohexol (A), plasma clearance of creatinine (B), or assessment of GFR/ECFV2comp (C) of iohexol in 51 of the 57 healthy cats in Figure 1. Six cats that were azotemic, Birman cats, or both were excluded from these analyses. Clearance (CLtrap) was calculated by use of the straight line as defined by data from 4 plasma samples obtained after IV injection of the markers that were converted to trapezoidal method estimates by use of a regression formula. Solid line represents regression line. See Figure 1 for key.
Citation: American Journal of Veterinary Research 70, 2; 10.2460/ajvr.70.2.176
Relationship between kilograms of body weight and estimated GFR determined on the basis of plasma clearance of iohexol (A), plasma clearance of creatinine (B), or assessment of GFR/ECFV2comp (C) of iohexol in 51 of the 57 study cats in Figure 1. Six cats that were azotemic, Birman cats, or both were excluded from these analyses. Solid line represents the regression line. See Figures 1 and 2 for key.
Citation: American Journal of Veterinary Research 70, 2; 10.2460/ajvr.70.2.176
Relationship between plasma creatinine concentration and estimated GFR determined on the basis of plasma clearance of iohexol (A) or plasma clearance of creatinine (B) in 57 cats. See Figures 1 and 2 for key.
Citation: American Journal of Veterinary Research 70, 2; 10.2460/ajvr.70.2.176
In most cats, the GFR estimates predicted by use of the dog-specific correction formula were closely related to the estimates predicted by use of the Brøchner-Mortensen formula. Only for cats with high CLtrap did the difference between estimates derived from the 2 formulae exceed 5%. The differences were 25% to 40% among the 4 cats with the highest clearance values.
The effect of reduction in number of samples from 4 to 2 on accuracy was evaluated. The loss of accuracy was small when the 2- and 3-hour samples were used for iohexol determinations and the 3- and 5-hour samples were used for creatinine determinations (Figure 1; Table 1). However, when the time points used for the 2-sample calculations were switched (ie, the 3- and 5-hour samples were used for iohexol determinations and the 2- and 3-hour samples were used for creatinine determinations), the 2-sample clearance deviated from the 4-sample clearance by a mean value of 2% for iohexol and 32% for creatinine.
Discussion
The difference in GFR values obtained by use of the 2 markers (mean difference, 13%) in the cats of the present study was of the magnitude commonly observed in method comparison studies and highlights the need to use method-specific reference values.
Commonly, agreement analysis is used to evaluate whether 2 methods yield identical results (ie, have high agreement). For the 2 GFR markers used in the cats of the present study, agreement analysis revealed that the corrected (CLtrap) plasma clearance methods can be used interchangeably (despite a 13% difference in measurements), provided that the reference range for the marker in question is used when evaluating the results. The differences were evenly spread around the mean difference, without systematic changes related to the level of GFR, as indicated by the high agreement index. This is, as expected, not as evident for the uncorrected (CL1comp) values. The use of the uncorrected results generates errors because of the part of the AUC that is not used in the calculation. The error generated differs with varying renal function; the error is large when clearance is high and negligible when clearance is low,7 thereby creating a systematic and changing bias related to the level of GFR. Thus, the agreement index for comparison of uncorrected clearance values is typically somewhat lower than that for corrected clearance values.
In the present study, agreement analysis also revealed that the use of 2 samples rather than 4 samples provided nearly identical results for both markers. Whereas iohexol distributes in the ECFV, creatinine distributes in total body water; therefore, the halflife of creatinine is approximately 3 times as long as that of iohexol. Although most of the AUC is defined at 5 hours after injection for iohexol, the excretion of creatinine takes 3 times longer. Such a longer excretion time was observed in our study, wherein the proportional reduction of plasma concentration of iohexol from 2 to 5 hours after injection differed from the proportional reduction of plasma concentration of creatinine during that same interval. For that reason, the 2 samples chosen for use were those obtained at 2 and 3 hours for iohexol and at 3 and 5 hours for creatinine. When the samples obtained at 3 and 5 hours and at 2 and 3 hours were used for iohexol and creatinine, respectively, the deviation from the 4-sample clearance was essentially unchanged for iohexol, whereas the deviation from the 4-sample clearance increased to 32% for creatinine. This illustrates the problem of collecting samples before the terminal monoexponential slope is reached. Results of a study32 in dogs with apparently normal or moderately reduced renal function indicated no benefit of sample collection at 2, 3, and 4 hours versus sample collection at only 2 and 3 hours with regard to iohexol assessments. Consequently, we used samples collected at 2 and 3 hours for the 2-sample method in our study, although sample collection until 5 hours is necessary for creatinine assessments. Even at 5 hours, only 62% of the total AUC for creatinine was defined. A reduction to 2 samples for assessment of creatinine may be unsuitable if renal function is compromised. If the 2-sample result for creatinine is low, the use of 4 samples may improve accuracy. Future research should aim to address problems of the limited-sample methods in animals with reduced renal function.
The option to collect only 2 blood samples has advantages because collection of many blood samples from nonsedated cats is sometimes difficult. If the deviation from the 4-sample result is clinically relevant in < 2% of cats, the veterinarian may choose to use the 2-sample method, particularly in instances where the alternative is no estimation of GFR in a given animal.
The regression analysis underlying the agreement analysis in this type of study is problematic because the degree of variation in the substance of interest is not known (ie, no duplicate clearance measurements), but the variation is expected to be similar between the 2 methods. The plots and agreement analysis make it possible to evaluate how clinically important the differences between methods will be in various clinical or research settings.
A difference in estimated GFR between young and old cats has recently been reported,31 but such a finding was not evident in the present study. Substantial variation in estimated GFR exists among elderly humans40 and among dogs30 as well as in the cats in our study. Secondary damage to renal tissue during various pathologic processes during a lifetime may occur and result in wide individual variation in the GFR in otherwise healthy individuals. Although the hypothesis of reduced GFR in old animals and the findings of the previous study31 were not supported by the findings in the present study, the hypothesis can also not be rejected on the basis of our study results. Compared with that previous study,31 the effect of body size on results in the present study could have been more pronounced because we investigated a greater number of cats of greater breed variation. An effect of age may be apparent in most species, provided that a sufficiently large number of animals with a sufficiently wide range of ages are studied. Inclusion of a greater number of cats or a higher number of markedly old cats may have been necessary to reveal an age effect in a clinical population with substantial variation in breeds, age, and body size. Nevertheless, the data obtained in the present study were representative of the heterogeneous population of cats encountered in clinical practice.
The results of the study reported here indicated a small but significant effect of body size on estimated GFR and that this effect was comparatively more important than an effect of age. This was also observed in a recent study30 of 118 dogs that weighed 2.5 to 70 kg. Given the relatively small number of cats in the present study, we did not find it appropriate to calculate reference ranges for different weight groups. This, as well as a potential breed effect, should be investigated in future research. In view of a previous report41 of renal disease in the Birman breed and on the basis of the fact that 2 of the 4 Birmans in our study were severely azotemic, all Birmans were excluded from calculation of reference ranges, as were the 2 cats in which azotemia was detected despite a lack of clinical signs of disease. To further investigate the effects of age, body size, and other variables on the reference ranges for estimated GFR and serum creatinine concentration in cats, future studies should involve a substantially larger number of cats and perhaps determine that the study cats truly have normal renal function by means other than clinicopathologic assessments.
Although the lack of a cat-specific formula for GFR prediction from the limited-sample results represented a limitation of the present study, the magnitude of error would potentially be greater by use of uncorrected results. If GFR is low, the error may be small (ie, 5%), whereas if GFR is high, the error may be as much as 30%.7 Most cats in which GFR is measured are likely nonazotemic or mildly azotemic; thus, if the correction formula is not used, the error generated will potentially be significant. Additional research undertakings should aim to define cat-specific correction formulae from simplified methods.
When predicting the trapezoidal method clearance value from the 1-compartmental model in the present study, we used both a correction formula calculated from data derived from dogs with apparently normal or moderately reduced renal function32 and the Brøchner-Mortensen formula.26 In a recent study,30 these 2 formulae yielded highly similar GFR estimates in dogs of intermediate sizes, whereas the human-specific formula yielded large discrepancies in very small dogs. Although the dog-specific formula is calculated from dogs with a wide range of body sizes (including very small dogs), the human-derived formula is based on the relatively uniform body size of adult humans. The dog-specific formula may therefore be more applicable for use in cats. In the present study, cats with the highest clearance values generally had low body weights. The high discrepancy between these 2 formulae in miniature dogs,30 as well as in the cats of this study with high clearances, raises the question of whether the dog formula performs better in the cat populations than the human-specific formula. This discrepancy illustrates the need for research to define a cat-specific formula; until a cat-specific correction formula exists, the human-specific or dog-specific formula must be used.
Standardization of GFR to weight (in kg) was used primarily in the present study, although for iohexol data, standardization to BSA and ECFV was also performed. Because creatinine distributes in total body water rather than in the ECFV, the mathematical theory underlying the standardization to ECFV is not applicable for creatinine. Some authors argue that minimizing the spread in data could point to the optimal method of standardization to body size,12 whereas our opinion is that GFR, like body size or hair color, may have a natural variation within a species, and minimizing variation is not necessarily an aim in itself. A previous study32 in dogs revealed a quite systematic deviation toward relatively higher values in small dogs and relatively lower values in large dogs by use of standardization to BSA rather than to weight or ECFV. This raises the question of the validity of the commonly used BSA formulae and points to a requirement for well-designed research into the issue of data normalization before uniform recommendations can be made. In the present study in cats, standardization to BSA resulted in the largest coefficient of variation among all the standardization methods.
Plasma concentration is a secondary pharmacokinetic variable in that it is influenced by both clearance and volume of distribution. Clearance and volume of distribution are primary pharmacokinetic variables that are influenced only by the physiologic variation. Clearance of a marker reflects GFR to the extent that it fulfills the criteria for a filtration marker: it is eliminated only in the kidneys and only by filtration, without tubular secretion or reabsorption.7 In instances where single sample methods are used for clearance estimation, correction formulae make use of empirically derived formulae for estimation of an individual's volume of distribution.42 The 2-sample method for clearance estimation used in our study represents a mathematical model for prediction of the clearance value as determined by the complete plasma disappearance curve.
Among individuals, there is physiologic variability in clearance values. Because of its simplicity, plasma concentration of creatinine is often used to estimate GFR in human medicine, and formulae have been derived to increase the accuracy of GFR that is estimated on the basis of plasma concentration. Common prediction formulae are the Cockroft-Gault formula or the MDRD formula, the latter being the most widely used today. Many factors such as weight, gender, race, and serum and urine variables were taken into account during the calculation of the MDRD formula in a study43 of more than 1,000 patients. Such a large and comprehensive study has not yet been conducted in veterinary medicine, and such an undertaking may be complicated further by breed differences. Also, in human medicine, clearance studies are considered more accurate in some situations, such as in children.8
Endogenous creatinine production can also vary markedly from one cat to another. Two cats with the same basal serum creatinine concentration can have different creatinine clearances. When exogenous creatinine is used as a filtration marker, endogenous creatinine production is corrected for by subtracting the plasma creatinine value. A functional test such as clearance measurement is more accurate than assessment of plasma creatinine concentration for evaluation of a cat's renal function. This was revealed in a study14 where 2 cats with similar plasma creatinine clearances (2.5 and 2.3 mL/min/kg) had different basal plasma creatinine concentrations (139 and 221 μmol/L, respectively) because of markedly different endogenous creatinine production rates (57 and 83 mg/kg/d, respectively).
Differences in laboratories' reference ranges may be a result of methodological differences, but also attributable to the available healthy population from which the reference range is derived (eg, number, age, size, and breed diversity of animals used). In 1 recent study44 to compare serum creatinine to GFR measurement in dogs based on receiver operating characteristic curves, the authors recommended higher cutoff values for serum creatinine concentration for classification of dogs as azotemic than what are commonly applied in most laboratories. Based on this fact, for the calculation of reference ranges, we included 2 cats with plasma creatinine concentrations that were 1% to 5% greater than the laboratory's reference range.
Although the present study provided reference ranges based on data obtained from a larger number of cats than the groups used in most previously published studies9–21 in dogs and cats, the limitations of these reference intervals should be acknowledged. Although owners perceived their cats to be healthy and this was confirmed by results of clinical examination and routine clinicopathologic analyses, it is possible that some cats did have occult renal disease. Urinalysis, abdominal ultrasonography, or other simple means of evaluating the kidneys should be included in future studies of reference ranges. In our study, the wide reference ranges determined in a relatively small number of cats produced relatively wide 95% confidence intervals for the upper and lower limit of the reference ranges.
Nevertheless, estimation of GFR will often help the clinician in cats with so-called borderline plasma creatinine values (ie, values that are slightly greater or less than the reference range of the laboratory [in our study, 180 μmol/L]). Among the 6 cats with borderline plasma creatinine concentrations in our study, some had clearance values at the lower end of the reference range (ie, clearances < 1.5 mL/min/kg), whereas others had high clearance values (ie, clearances > 2.5 mL/min/kg). For the former individuals, there is indication to further investigate and monitor the cats with respect to kidney disease; for the latter individuals, the data do not support compromised kidney function in those cats.
No adverse drug reactions were observed in our investigation in cats or in other studies10,12,14,17,19,32,45–50 in which iohexol was used for GFR estimation in cats and dogs. With the extensive and increasing use of radiographic contrast media in all areas of medicine, the prospect of adverse drug reactions is a concern. Although data on adverse drug reactions in animals is scarce, adverse reactions were uncommon in a large prospective study51 of more than 300,000 human patients receiving radiographic contrast media, with a calculated prevalence rate of severe adverse effects of 0.04%. Administration of high doses of iohexol for radiographic contrast studies also appears quite safe even in high-risk patients that have markedly reduced renal function, although transient adverse drug reactions have been observed.52,53 The effect of radiographic contrast media on renal function in humans has been investigated quite extensively.23 In a large prospective study54 to investigate immediate and delayed adverse drug reactions after administration of iohexol, minor and transient adverse drug reactions (eg, nausea) were detected in as many as 3% of patients, but not a single severe or life-threatening adverse drug reaction occurred. In patients in an intensive care unit, some with severely reduced renal function, GFR estimation by use of iohexol was performed without adverse drug reactions.55 In contrast, severe adverse drug reactions have been described with the use of inulin.56,57
On the basis of results of our study, the 4-sample method provides a reliable and convenient means of estimating GFR in cats with GFRs that are considered normal or near normal. The 2-sample method also provides acceptable approximations for GFR determined by measurement of plasma clearance of iohexol or creatinine; via this method, the GFR estimation procedure is clinically more applicable in nonsedated animals. However, compared with the 2-sample method, the 4-sample method provides data that are more accurate in cats with low renal function. When interpreting estimated GFR values, the body size of the cat should be taken into account.
ABBREVIATIONS
| AUC | Area under the curve |
| BSA | Body surface area |
| CL1comp | Clearance as measured by use of a 1- compartmental model |
| CLtrap | Clearance as measured by use of a noncompartmental (trapezoidal) model |
| ECFV | Extracellular fluid volume |
| GFR | Glomerular filtration rate |
| GFR/ECFV1comp | Glomerular filtration rate standardized to liters of extracellular fluid volume by use of a 1-compartmental model |
| GFR/ECFV2comp | Glomerular filtration rate standardized to liters of extracellular fluid volume by use of a 2-compartmental model |
| MDRD | Modification of diet in renal disease |
Central Laboratory, Norwegian School of Veterinary Science, Oslo, Norway.
Omnipaque, Nycomed Amersham General Electric, Oslo, Norway.
Vitros 250 chemistry system, Ortho-Clinical Diagnostics, Raritan, NJ.
Performance Verifier I and II control sera, Ortho-Clinical Diagnostics, Issy-lès-Moulineaux, France.
WinNonlin, version 5.2, Pharsight Corp, Mountain View, Calif.
Excel 2003, Microsoft Corp, Redmond, Wash.
References
- 1.
Elliott J, Barber PJ. Feline chronic renal failure: clinical findings in 80 cases diagnosed between 1992 and 1995. J Small Anim Pract 1998;39:78–85.
- 2.↑
Elliott J, Rawlings JM, Markwell PJ, et al. Survival of cats with naturally occurring chronic renal failure: effect of dietary management. J Small Anim Pract 2000;41:235–242.
- 3.
Kuwahara Y, Ohba Y, Kitoh K, et al. Association of laboratory data and death within one month in cats with chronic renal failure. J Small Anim Pract 2006;47:446–450.
- 4.
King JN, Gunn-Moore DA, Tasker S, et al. Tolerability and efficacy of benazepril in cats with chronic kidney disease. J Vet Intern Med 2006;20:1054–1064.
- 5.
Mizutani H, Koyama H, Watanabe T, et al. Evaluation of the clinical efficacy of benazepril in the treatment of chronic renal insufficiency in cats. J Vet Intern Med 2006;20:1074–1079.
- 6.↑
Bailey DB, Rassnick KM, Erb HN, et al. Effect of glomerular filtration rate on clearance and myelotoxicity of carboplatin in cats with tumors. Am J Vet Res 2004;65:1502–1507.
- 7.↑
Heiene R, Moe L. Pharmacokinetic aspects of measurement of glomerular filtration rate in the dog: a review. J Vet Intern Med 1998;12:401–414.
- 8.↑
Hjorth L, Wiebe T, Karpman D. Correct evaluation of renal glomerular filtration rate requires clearance assays. Pediatr Nephrol 2002;17:847–851.
- 9.
Barthez PY, Chew DJ, DiBartola SP. Effect of sample number and time on determination of plasma clearance of technetium Tc 99m pentetate and orthoiodohippurate sodium I 131 in dogs and cats. Am J Vet Res 2000;61:280–285.
- 10.↑
Brown SA, Finco DR, Boudinot FD, et al. Evaluation of a single injection method, using iohexol, for estimating glomerular filtration rate in cats and dogs. Am J Vet Res 1996;57:105–110.
- 11.
Brown SA, Haberman C, Finco DR. Use of plasma clearance of inulin for estimating glomerular filtration rate in cats. Am J Vet Res 1996;57:1702–1705.
- 12.↑
Goy-Thollot I, Chafotte C, Besse S. Iohexol plasma clearance in healthy dogs and cats. Vet Radiol Ultrasound 2006;47:168–173.
- 13.
Haller M, Rohner K, Muller W, et al. Single-injection inulin clearance for routine measurement of glomerular filtration rate in cats. J Feline Med Surg 2003;5:175–181.
- 14.↑
Le Garreres A, Laroute V, De La Farge F. Disposition of plasma creatinine in non-azotaemic and moderately azotaemic cats. J Feline Med Surg 2007;9:89–96.
- 15.
McClellan JM, Goldstein RE, Erb HN, et al. Effects of administration of fluids and diuretics on glomerular filtration rate, renal blood flow, and urine output in healthy awake cats. Am J Vet Res 2006;67:715–722.
- 16.
Miyamoto K. Evaluation of single-injection method of inulin and creatinine as a renal function test in normal cats. J Vet Med Sci 1998;60:327–332.
- 17.
Miyamoto K. Clinical application of plasma clearance of iohexol on feline patients. J Feline Med Surg 2001;3:143–147.
- 18.
Miyamoto K. Evaluation of plasma clearance of inulin in clinically normal and partially nephrectomized cats. Am J Vet Res 2001;62:1332–1335.
- 19.
Miyamoto K. Use of plasma clearance of iohexol for estimating glomerular filtration rate in cats. Am J Vet Res 2001;62:572–575.
- 20.
Rogers KS, Komkov A, Brown SA, et al. Comparison of four methods of estimating glomerular filtration rate in cats (Erratum published in Am J Vet Res 1991;52:1742). Am J Vet Res 1991;52:961–964.
- 21.
Uribe D, Krawiec DR, Twardock AR, et al. Quantitative renal scintigraphic determination of the glomerular filtration rate in cats with normal and abnormal kidney function, using 99mTc-diethylenetriaminepentaacetic acid. Am J Vet Res 1992;53:1101–1107.
- 22.
Brown SCW, O'Reilly PH. Iohexol clearance for the determination of glomerular filtration rate in clinical practice: evidence for a new gold standard. J Urol 1991;146:675–679.
- 24.↑
Shemesh O, Golbetz H, Kriss JP. Limitations of creatinine as a filtration marker in glomerulopathic patients. Kidney Int 1985;28:830–838.
- 25.
Watson AD, Lefebvre HP, Concordet D. Plasma exogenous creatinine clearance test in dogs: comparison with other methods and proposed limited sampling strategy. J Vet Intern Med 2002;16:22–33.
- 26.↑
Brøchner-Mortensen J. A simple method for the determination of glomerular filtration rate. Scand J Clin Lab Invest 1972;30:271–274.
- 27.
Hoang K, Tan JC, Derby G. Determinants of glomerular hypofiltration in aging humans. Kidney Int 2003;64:1417–1424.
- 28.
Pomeroy MJ, Robertson JL. The relationship of age, sex, and glomerular location to the development of spontaneous lesions in the canine kidney: analysis of a life-span study. Toxicol Pathol 2004;32:237–242.
- 30.↑
Bexfield NH, Heiene R, Gerritsen RJ. Glomerular filtration rate estimated by 3-sample plasma clearance of iohexol in 118 healthy dogs. J Vet Intern Med 2008;22:66–73.
- 31.↑
van Hoek I, Vandermeulen E, Duchateau L. Comparison and reproducibility of plasma clearance of exogenous creatinine, exo-iohexol, endo-iohexol, and 51Cr-EDTA in young adult and aged healthy cats. J Vet Intern Med 2007;21:950–958.
- 32.↑
Heiene R, Moe L. The relationship between some plasma clearance methods for estimation of glomerular filtration rate in dogs with pyometra. J Vet Intern Med 1999;13:587–596.
- 34.
Altman DG, Bland JM. Measurement in medicine: the analysis of method comparison studies. Statistician 1983;32:307–317.
- 35.
Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–310.
- 36.↑
Altman DG. Statistics in medical journals: developments in the 1980s. Stat Med 1991;10:1897–1913.
- 37.
Aaras A, Veierod MB, Larsen S. Reproducibility and stability of normalized EMG measurements on musculus trapezius. Ergonomics 1996;39:171–185.
- 38.
Veierød MB, Larsen S. Definitions, properties and applications of the stability ratio. Statistics in medicine, in Proceedings. Int Soc Clin Biostat Cong 1992;72.
- 39.↑
Nordic Committee on Food Analysis (NMKL). New procedure: NMKL procedure No. 20, 2007: evaluation of results from qualitative methods. NMKL Newsletter 2007;65:1.
- 40.↑
Iseki K, Ikemiya Y, Iseki C, et al. Proteinuria and the risk of developing end-stage renal disease. Kidney Int 2003;63:1468–1474.
- 41.↑
Gunn-Moore DA, Dodkin SJ, Sparkes AH. An unexpectedly high prevalence of azotaemia in Birman cats. J Feline Med Surg 2002;4:165–166.
- 42.↑
Stake G, Monclair T. A single plasma sample method for estimation of the glomerular filtration rate in infants and children using iohexol, I: establishment of a body weight-related formula for the distribution volume of iohexol. Scand J Clin Lab Invest 1991;51:335–342.
- 43.↑
Levey AS, Bosch JP, Lewis JB, et al. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of diet in Renal Disease Study Group. Ann Intern Med 1999;130:461–470.
- 44.↑
Hartmann H, Mohr S, Thüre S, et al. Routine use of a renal function test for quantitative assessment of glomerular filtration rate (GFR) including determination of a cut-off value for azotemia in the dog. Eur J Companion Anim Pract 2008;18:29–36.
- 45.
Braselton WE, Stuart KJ, Kruger JM. Measurement of serum iohexol by determination of iodine with inductively coupled plasma-atomic emission spectroscopy. Clin Chem 1997;43:1429–1435.
- 46.
Finco DR, Braselton WE, Cooper TA. Relationship between plasma iohexol clearance and urinary exogenous creatinine clearance in dogs. J Vet Intern Med 2001;15:368–373.
- 47.
Gleadhill A, Michell AR. Evaluation of iohexol as a marker for the clinical measurement of glomerular filtration rate in dogs. Res Vet Sci 1996;60:117–121.
- 48.
Laroute V, Lefebvre HP, Costes G, et al. Measurement of glomerular filtration rate and effective renal plasma flow in the conscious beagle dog by single intravenous bolus of iohexol and p-aminohippuric acid. J Pharmacol Toxicol Methods 1999;41:17–25.
- 49.
Laroute V, Chetboul V, Roche L, et al. Quantitative evaluation of renal function in healthy Beagle puppies and mature dogs. Res Vet Sci 2005;79:161–167.
- 50.
Moe L, Heiene R. Estimation of glomerular filtration rate in dogs with 99mTc-DTPA and iohexol. Res Vet Sci 1995;58:138–143.
- 51.↑
Katayama H, Yamaguchi K, Takashima T, et al. Adverse reactions to ionic and nonionic contrast media: a report from the Japanese committee on the safety of contrast media. Radiology 1990;175:621–628.
- 52.
Rudnick MR, Goldfarb S, Wexler L, et al. Nephrotoxicity of ionic and nonionic contrast media in 1196 patients: a randomized trial. Kidney Int 1995;47:254–261.
- 53.
Schwartz GJ, Furth S, Cole SR, et al. Glomerular filtration rate via plasma iohexol disappearance: pilot study for chronic kidney disease in children. Kidney Int 2006;69:2070–2077.
- 54.↑
Munechika H, Hiramatsu Y, Kudo S, et al. A prospective survey of delayed adverse reactions to iohexol in urography and computed tomography. Eur Radiol 2003;13:185–194.
- 55.↑
Erley CM, Bader BD, Berger ED, et al. Plasma clearance of iodine contrast media as a measure of glomerular filtration rate in critically ill patients. Crit Care Med 2001;29:1544–1550.
- 56.
Chandra R, Barron JL. Anaphylactic reaction to intravenous sinistrin (Inutest). Ann Clin Biochem 2002;39:76.
- 57.
Franck P, Moneret-Vautrin DA, Morisset M, et al. Anaphylactic reaction to inulin: first identification of specific IgEs to an inulin protein compound. Int Arch Allergy Immunol 2005;136:155–158.
