Glomerular filtration rate in dogs and cats

Joanna White

 Consolidated from IRIS educational articles: GFR in Practice: Assessment of Glomerular Filtration Rate in Dogs (updated by R Heiene from Heiene, Lefebvre & Watson, 2015) and Glomerular Filtration Rate in Dogs and Cats (Heiene & Lefebvre)

Dedication

Introduction

Glomerular filtration rate (GFR) is considered one of the most useful and sensitive indicators of overall renal function, although it has limitations. Due to renal functional reserve and compensatory hyperfiltration in remaining nephrons, GFR can remain within the normal range, despite substantial underlying renal pathology, and should therefore be interpreted alongside other clinical findings rather than in isolation. Any decrease in GFR, when other pre-renal causes have been excluded, generally signifies that kidney disease is occurring.

In most patients, clinical evaluation — history, physical examination, urine specific gravity and protein content, blood creatinine, and renal imaging — is sufficient to determine whether overall kidney function is compromised. However, certain presentations warrant further assessment: for example animals with polyuria/polydipsia where other diagnoses have been excluded, non-azotemic patients where subclinical renal disease is suspected.

The IRIS staging system is currently based on blood creatinine and SDMA concentrations, but the IRIS board considers GFR measurement a potential future primary staging criterion, as is already the standard in human medicine. Several GFR measurement methods have been validated in dogs and cats, with plasma clearance techniques using non-radioactive markers — particularly creatinine and iohexol — offering practical alternatives to urinary clearance methods in both clinical and research settings.


Blood Sampling Strategy

Available Markers

Principle of plasma clearance

Physiological Factors Affecting GFR

GFR is estimated by measuring the plasma clearance (Cl) of an exogenous marker. Clearance is calculated as:

  GFR = Cl = Dose / AUC

 where AUC is the area under the plasma concentration-versus-time curve following intravenous administration of the marker. AUC is determined using a 1- or 2-compartment model or by the trapezoidal method. A suitable marker must be:

·       Freely filtered by the glomerulus

·       Not subject to tubular resorption or secretion

·       Neither metabolised or eliminated by extrarenal means

 Both iohexol and exogenous creatinine fulfil these criteria in dogs and cats.

Fig 1: Example of a plasma disappearance curve, where the elimination is monoexponential after an initial "distribution phase" of the marker

The greater the number of samples and the longer the collection period, the more accurate the AUC estimation and therefore the GFR. However, practical limitations require limited-sampling strategies in dogs and cats. The timing of blood sampling should minimise the proportion of AUC that is extrapolated beyond the last measured time point. Ideally, the extrapolated area should be less than 20% of the total AUC. As renal function declines, marker elimination slows, making later sampling increasingly important when renal impairment is suspected.

Accuracy of dose administration and precise recording of exact sampling times are critical — errors in either will produce erroneous GFR calculations. If a sample is collected late, record the actual collection time (e.g. if the 120-minute sample is collected 4 minutes late, record 124 minutes).

 Repeated collection of 5 mL blood samples in cats or miniature dogs may be difficult and risk excessive blood loss. Assays requiring only small plasma volumes (e.g. 0.2 mL for enzymatic creatinine assay) reduce total blood withdrawal. Paediatric capillary-based sampling devices have been proposed as alternatives, offering advantages including absence of vein collapse, limited blood withdrawal and improved safety.

 Correction formulae derived from human medicine (e.g. the Brøchner-Mortensen formula) have been adapted for dogs and cats to estimate actual GFR from limited-sampling strategies which tend to focus on the elimination phase and ignore the contribution of the redistribution phase to the AUC. Repeated clearance measurements in any individual patient should use the same technique to avoid misinterpretation from methodological differences, as differences of 10 to 20% can arise from inter-method variation alone.

Plasma Iohexol Clearance

Iohexol (300 mg iodine/mL) is a non-ionic radiographic contrast agent available as a medical-grade reagent. It is stable at room temperature, and samples can be mailed to referral laboratories without special precautions. Iohexol assays are commercially available through a limited number of centres; the laboratory will typically calculate the GFR from the dose, sampling times and measured concentrations provided by the clinician.

Materials Required

·       Iohexol solution (300 mg iodine/mL)

·       Sterile saline (0.9% NaCl)

·       Intravenous catheter with stopper; syringes, needles and blood sample tubes

·       Timer

 Procedure

·       Fast patient overnight, hospitalised, and providing water but not food.

·       Record current body weight (kg)

·       Administer iohexol solution as an accurate IV bolus: 1 mL/kg for non-azotaemic dogs

·       Flush catheter dead space with 2 mL saline, start timer immediately, then remove catheter

·       Collect blood samples (≥1 mL each) at 2, 3 and 4 hours post-administration, from sites different from the iohexol administration site, and record exact sampling times and label tubes accordingly

·       Centrifuge samples, harvest plasma, and send to laboratory

Cats: Dosing and Sampling

Published feline iohexol protocols have historically differed from the dog protocol above in both dose and sampling schedule. However, the commercial laboratories currently offering feline iohexol GFR testing (Royal Veterinary College/deltaDOT, UK; Michigan State University Veterinary Diagnostic Laboratory, USA) now use the same dose and sampling schedule in cats as in dogs, so a single protocol can be followed for both species.

Current standard protocol (RVC/deltaDOT and MSU VDL):

·       Patient fasted for approximately 12 hours and well hydrated

·       Administer iohexol 300 mg iodine/kg (1 mL/kg of 300 mg iodine/mL solution) as an accurate IV bolus, and record the exact start time

·       A baseline (time-zero) sample is not required. Collect blood samples at 2, 3 and 4 hours (120, 180 and 240 minutes) after injection, from a site other than that used for administration; harvest serum or plasma and submit to the laboratory with an accurately recorded body weight, dose given, and exact sampling times

·       Clearance is calculated from the 3-sample slope-intercept method using a 1-compartment model and corrected using a cat-specific formula (see below)

Older, more intensive feline protocols are described in the literature for context. Using X-ray fluorescence, Goy-Thollot et al. (2006) administered iohexol at 450 mg iodine/kg (1.5 mL/kg) as an IV bolus over 30–60 seconds, with 10 blood samples collected at 5, 20, 40, 60, 80, 100, 120, 150, 180 and 240 minutes after injection, and clearance calculated from a two-compartment model.

Using a colorimetric iodine assay, Miyamoto (2001) used a lower dose (90 mg iodine/kg in renal-intact cats; 45 mg iodine/kg in cats with reduced renal mass) with only 3 post-injection samples (120, 180 and 240 minutes in renal-intact cats; 120, 240 and 360 minutes in cats with reduced renal mass), using a modified 1-compartment model corrected by the Brochner-Mortensen formula.

Correction for Limited Sampling

A correction should be applied to the clearance value calculated from a limited (3-sample) protocol to improve accuracy, since the redistribution phase of the plasma disappearance curve is not directly sampled. The correction formula applied should be species-specific:

Dog (Heiene and Moe 1999):

Clcorr = −0.03 + 1.06(Clobs) − 0.07012(Clobs)²

Cats (Finch et al. 2011), derived and validated specifically for the 2-, 3- and 4-hour sampling protocol described above:

Clcorr = 1.036(Clobs) − 0.062(Clobs)²

where Clcorr = corrected clearance and Clobs = uncorrected clearance calculated from the 3-sample slope-intercept method.

In the validation study of 19 healthy cats, the feline-specific formula gave excellent agreement with reference 8-sample clearance, with the lowest mean percentage error (0.59 ± 7.05%) of the methods tested. The canine (Heiene) formula performed similarly well when applied to the same feline data (1.56 ± 6.95% error), whereas human Brochner-Mortensen correction formulae showed a non-linear relationship with reference clearance in cats and are not recommended for this species (11.13% adult / 13.14% child mean error). Some laboratories apply the appropriate correction before reporting; confirm with the laboratory used which formula, if any, has been applied.

Advantages and Disadvantages

Advantages:

·       Stable marker; samples can be mailed without special precautions

·       Relatively short total procedure time (4 hours)

·       Laboratory calculates GFR from clinician-provided data

Disadvantages:

·       Samples must be sent to an external laboratory

·       Cost of iohexol may be limiting in large dogs

·       Adverse reactions have been reported (very rare in dogs)

·       Correction formula required for 3-sample strategy

Commercial Laboratories

Contact the laboratory beforehand to confirm pricing, sample requirements and whether the correction formula has been applied to the reference range provided:

· The Royal Veterinary College, UK, in collaboration with deltaDOT (www.rvc.ac.uk/pathology-and-diagnostic-laboratories/therapeutic-drug-monitoring). The samples are tested using high performance capillary electrophoresis. Dogs and cats use the same protocol: iohexol 300 mg iodine/kg IV, with samples at 2, 3 and 4 hours; feline results are corrected using the Finch et al. (2011) formula.

-   Michigan State University Veterinary Diagnostic Laboratory (cvm.msu.edu/vdl). Dogs and cats use the same protocol: iohexol 300 mg iodine/kg IV, with samples at 2, 3 and 4 hours; iohexol is measured by inductively coupled plasma mass spectrometry and reported as clearance and percentage reduction relative to a normal cohort.

Plasma Creatinine Clearance

Exogenous creatinine (anhydrous creatinine powder, e.g. Sigma Aldrich) can be used as a GFR marker in dogs and cats. Although serum creatinine assays are widely available in-house, a sterile, pyrogen-free, precisely concentrated creatinine solution must be prepared by an appropriate laboratory or pharmacy before administration. The practitioner is responsible for calculating GFR from the data obtained.

Note: Creatinine has a 2–3-times longer elimination half-life than most other GFR markers due to a larger volume of distribution. This requires a substantially longer hospitalisation period than iohexol. Blood sampling at 1 and 10 hours post-administration provides a relatively accurate AUC estimation (maximum error −14% vs. an 11-point kinetic profile), but sampling at inappropriate times can alter results significantly. Overnight hospitalisation for a 24-hour sample may be required in some patients.

Materials Required

·       Anhydrous creatinine powder (Sigma Aldrich)

·       Sterile saline (0.9% NaCl)

·       0.2 μm filter (via pharmacy)

·       Intravenous catheter with stopper; syringes, needles and blood sample tubes

·       Timer

Procedure

·       Fast patient overnight, hospitalised, and providing water but not food

·       Record current body weight (kg)

·       Collect a baseline blood sample (time zero) to measure pre-injection creatinine concentration

·       Prepare creatinine solution (80 mg/mL): dissolve in saline and sterilise by filtration through a 0.2 μm filter

·       Administer creatinine solution (1 mL/kg) as an accurate IV bolus, followed by 2 mL saline. Start timer immediately. Remove catheter

·       Collect 1 mL blood samples at 10 minutes and at 1, 2, 6 and 10 hours post-administration, recording exact sampling times. Use sites different from the administration site

·       Analyse creatinine concentrations in all 6 samples (baseline + 5 post-injection) in the same assay batch

·       Calculate plasma clearance from dose administered and AUC, after subtracting the baseline value from each post-injection concentration

Advantages and Disadvantages

Advantages:

·       Creatinine assays widely available in-practice or externally

·       Small required sample volume (0.2 mL sufficient for enzymatic assay)

·       Short assay turnaround time

Disadvantages:

·       Sterile creatinine solution must be prepared by an approved laboratory or pharmacy

·       Substantially longer hospitalisation required than with iohexol

·       GFR calculation is performed by the clinician, not an external laboratory


Reproducibility of GFR Measurements

When repeating GFR measurements in an individual patient, it is important to distinguish clinically relevant change from physiological and analytical variability. Between-day coefficients of variation are generally less than 20% for most validated markers in both healthy and diseased animals, which is considered acceptable. To avoid misinterpretation from methodological differences, all repeated measurements in each patient should use the same technique, as inter-method differences of 10–20% are commonly observed even when studies are conducted in the same animals at the same laboratory.


Indexation of glomerular filtration rate

Clearance is measured as mL/min. Standardization (indexation; scaling) of GFR values to the body weight allows comparison of the clearance value to reference ranges for healthy animals. GFR in dogs and cats is traditionally expressed in mL/min/kg. A current challenge is to define the most appropriate way to standardize GFR. Body surface area (BSA) has been proposed as the reference for indexing physiological variables and it is routinely used in human medicine. However, it has been recommended that such an indexation for GFR should be abandonned²⁸ and the formulae used to estimate BSA in dogs are probably inaccurate.²⁹ Indexation to extracellular fluid volume (ECFV) has been proposed as an alternative because one of the major roles of the kidney is to regulate body fluid composition.³⁰ Nevertheless, standardization to ECFV did not produce substantial changes in the relationships between GFR estimates and body weight in adult dogs.⁵ Moreover, with indexation to ECFV, differences between puppies and adult dogs were still observed, but were inversed.²⁰ In an earlier canine study, standardization to body weight, BSA or ECFV was shown to produce quite different results and for some dogs altered the clinical interpretation of the GFR value obtained.¹⁰ A similar problem exists in the cat: standardization to BSA resulted in larger between-individual coefficient of variation (36%) than did standardization to body weight (27%) or ECFV (24%).⁹

Standardization of GFR is also of concern in obese human patients, who are clearly different from lean individuals of similar body weights. The higher the weight, the lower the GFR indexed to BSA. Other ways to index GFR have been tested unsuccessfully. An absolute, non-corrected GFR is currently recommended in obese patients.³¹ There are no published data regarding indexation of body weight in obese dogs or cats. In children, variation in body shape or constitution causes difficulties with prediction formulae and simplified approaches.³²

Further research is needed in order to evaluate the best method for standardization of GFR values in dogs and cats. Such research is not simple because there is no "golden standard" to relate to. Ideally, regression analysis should be used on a large representative population of adult dogs (different breeds, body weights, age and sex), evaluating different methods of standardization. A simplified alternative would be to stratify the canine population into body weight categories and define an “average” cut-off for the GFR estimates in each category. The main limitation then would be that the greater the range of body weights for a given category, the more inaccurate the corresponding derived cut-off values. For the time being, it seems logical to standardize to body weight as most publications on reference intervals make use of this approach.


Reference intervals

Dogs                                      

Most published GFR values in dogs fall between approximately 2 and 5 mL/min/kg. A value below 1.5 mL/min/kg can be considered abnormal (Heiene and Lefebvre 2007), but a single universal cut-off is not appropriate given the body weight dependency of GFR.

 

In a study of 113 healthy dogs stratified by body weight category (Mini, Medium, Maxi, Giant), mean (± SD) GFR values were 3.7 ± 0.5, 3.0 ± 0.5, 2.5 ± 0.4 and 2.4 ± 0.6 mL/min/kg respectively, with corresponding approximate lower reference limits of 2.7, 2.0, 1.7 and 1.3 mL/min/kg. In a second study of 118 healthy adult dogs, reference ranges by body weight quartile were 1.54–4.25, 1.29–3.50, 1.20–3.36 and 1.12–3.39 mL/min/kg for weight quartiles of 1.8–12.4, 13.2–25.5, 25.7–31.6 and 32.0–70.3 kg respectively. These data demonstrate that using a single cut-off of 1.5 mL/min/kg will yield false-negative results in small dogs and false-positive results in giant dogs.

 

Breed-specific tentative reference intervals have been reported: English Pointers 2.3–5.1 mL/min/kg, English Setters 1.8–5.0 mL/min/kg, German Shepherds 1.7–3.8 mL/min/kg. In practice, most breeds tend to follow their body weight category for reference values. Stratification by body weight is recommended; age stratification does not appear necessary in adult animals.

 

Cats

Proposed plasma clearance reference ranges for adult cats are 1.0–3.5 mL/min/kg for iohexol and 1.3–3.8 mL/min/kg for creatinine. Notably, the cut-off for GFR is lower in cats than in dogs of similar body weight, which is consistent with allometric relationships between body weight and renal function across species. These preliminary reference intervals require validation in larger feline populations.

 

Unlike in dogs, normalisation to body weight, body surface area, or extracellular fluid volume all performed comparably well in cats, likely reflecting the narrower range of body sizes within the species, so any of the three methods can reasonably be used clinically (Goy-Thollot et al. 2006)

 


GFR in Non-Primary Renal Diseases

Summary Table: Reported GFR Reference Ranges

This article is dedicated to memory of Dr Reidun Heiene, whose work laid much of the foundation for our understanding of glomerular filtration rate assessment in dogs and cats. Across three decades of research into plasma clearance methodology — from the pharmacokinetic principles of iohexol clearance to the practical correction formulae that make limited-sampling protocols possible in clinical practice — Dr Heiene's contributions shaped the field in ways that continue to guide veterinary nephrology today.

GFR assessment may be valuable in evaluating patients with concurrent cardiovascular or endocrine diseases, where renal function may be compromised but azotaemia is absent or only mild.

 Cardiovascular Disease

Azotaemia is common in dogs with cardiac disease but is typically mild to moderate. In one study of 24 dogs with chronic valvular disease, GFR was significantly lower in NYHA class III–IV (1.7 ± 0.7 mL/min/kg) than in class I–II (3.1 ± 0.8 mL/min/kg), and the degree of azotaemia did not correlate with cardiovascular variables. Only 1/15 NYHA I–II dogs had a GFR below 2 mL/min/kg, compared with 7/9 NYHA III–IV dogs. Further investigation is needed to identify the cause and clinical relevance of reduced GFR in canine cardiac disease.

 Endocrine Disease

Feline Hyperthyroidism

Hyperthyroidism is common in older cats, and a 14–40% prevalence of pre-existing CKD has been reported in affected individuals. Hyperthyroidism can mask underlying CKD by increasing GFR, making renal function assessment critical before and after treatment. CKD becomes clinically apparent after treatment in approximately 17–39% of hyperthyroid cats. Pre-treatment GFR has been shown to predict the risk of post-treatment azotaemia. An accurate evaluation of kidney function can be performed from 4 weeks after radioiodine treatment.

 Canine Hypothyroidism

GFR is reduced in experimentally induced canine hypothyroidism, while blood creatinine concentrations remain unaltered. In a series of 14 dogs with naturally occurring hypothyroidism, GFR was below 2 mL/min/kg in all cases and improved with levothyroxine treatment. The clinical relevance of renal dysfunction and subclinical CKD in hypothyroid dogs requires further investigation.

 Other Endocrine Conditions

The utility of GFR estimation in diabetes mellitus, hypoadrenocorticism, spontaneous hyperadrenocorticism or during glucocorticoid therapy is currently limited by insufficient data on how these conditions affect GFR; further studies are needed