Abstract
OBJECTIVE
To evaluate glomerular filtration rate (GFR) estimation by means of plasma clearance of iohexol (IOX) in domestic rabbits and to assess accuracy of limited-sampling models for GFR estimation.
ANIMALS
6 healthy domestic rabbits (Oryctolagus cuniculus).
PROCEDURES
Each rabbit received IOX (64.7 mg/kg [0.1 mL/kg], IV), and blood samples were collected at predetermined times before and after administration. Plasma IOX concentration was determined by high-performance liquid chromatography. The pharmacokinetics of IOX was determined by a noncompartmental method. For each rabbit, plasma clearance of IOX was determined by dividing the total IOX dose administered by the area under the concentration-time curve indexed to the subject's body weight. The GFR estimated from the plasma IOX concentration at 6 sampling times (referent model) was compared with that estimated from the plasma IOX concentration at 5 (model A), 4 (model B), and 3 (models C, D, and E) sampling times (limited-sampling models).
RESULTS
Mean ± SD GFR was 4.41 ± 1.10 mL/min/kg for the referent model and did not differ significantly from the GFR estimated by any of the limited-sampling models. The GFR bias magnitude relative to the referent model was smallest for model D in which GFR was estimated from plasma IOX concentrations at 5, 15, and 90 minutes after IOX administration.
CONCLUSIONS AND CLINICAL RELEVANCE
Results suggested that plasma clearance of IOX was a safe, reliable, accurate, and clinically feasible method to estimate GFR in domestic rabbits. Further research is necessary to refine the method.
Chronic kidney disease is common in domestic rabbits (Oryctolagus cuniculus) and is often difficult to diagnose, especially during the early stages of the disease. In rabbits, nephrons are intermittently operative, and nephrons are never all active at any 1 time. When there is an incipient reduction in functional renal mass, the remaining glomeruli of dogs and cats enter a state of hyperfiltration to compensate, whereas rabbits recruit silent (inactive) glomeruli. Consequently, in rabbits, the early stages of CKD are characterized by the activation of previously inactive nephrons, which masks the signs of impaired renal function. In fact, rabbits with CKD can live for months to years without any clinical signs of the disease. However, once rabbits lose > 50% to 75% of renal function, they have a significant increase in serum creatine and SUN concentrations in a manner similar to that of dogs and cats with impaired renal function. Although abnormally increased serum creatinine and SUN concentrations are commonly used to identify rabbits with CKD, those 2 biomarkers are not very sensitive for that purpose because they can be affected by several extrarenal factors or pathological conditions such as gut stasis.1
Glomerular filtration rate is considered the best quantitative variable for assessment of overall renal function in both human and veterinary medicine.2–5 It is directly related to functional renal mass and is considered the most sensitive biomarker for early-stage kidney disease.6,7 The GFR can be determined by urinary or plasma Cl of an ideal filtration marker, which is defined as a substance that is freely filtered by the kidney, not reabsorbed or secreted by the renal tubules, not metabolized or bound to plasma proteins, and not nephrotoxic.5,6 Iohexol is a nonionic low-osmolarity contrast medium that is widely used in many species because of its low toxicity. In fact, plasma Cl of IOX is used to estimate GFR in humans8 and various animals, such as horses,2 donkeys,2 dogs,6,9 and cats.7,9
To our knowledge, estimation of the GFR by means of IOX Cl has not been evaluated in rabbits. Therefore, the aims of the study reported here were to evaluate GFR estimation by means of plasma IOX Cl in healthy domestic rabbits and to assess the accuracy of limited-sampling methods for GFR estimation.
Materials and Methods
Animals
The study was conducted in accordance with the guidelines established by European law regarding the use of animals in research, and all study procedures were reviewed and approved by the Pisa University Animal Ethics and Welfare Committee (No. 15910/2017). Privately owned animals were considered for study enrollment following the acquisition of consent from the owners. From each potential study subject, 1 mL of blood was collected from a marginal ear vein and a urine sample (1 mL) was collected by free catch. All rabbits were screened for antibodies against Encephalitozoon cuniculi by means of a commercially available ELISA.a Rabbits that were seropositive for antibodies against E cuniculi were excluded from the study. Only rabbits that were considered healthy on the basis of results of a physical examination, hematologic and serum biochemical analyses, urinalysis, and GGT index (ratio of urine GGT activity to urine creatinine concentration) were enrolled in the study. This selection process resulted in a study population of 6 rabbits.
Sample collection and analysis
Rabbits remained conscious and unsedated during collection of all samples and IOX administration and had ad libitum access to food and water. From each rabbit, a blood sample (2.5 mL) was collected by jugular venipuncture; 1.0 mL was collected into a blood collection tube containing potassium EDTA (EDTA tube) as an anticoagulant, and the remaining 1.5 mL was collected into a sterile blood collection tube without any additives or anticoagulant. The blood collected into the EDTA tube underwent a CBC with an automated laser analyzerb and a manual differential WBC count. For the differential WBC count, a blood smear was created and stained with May-Grünwald-Giemsa stain, then evaluated with a microscope. An automated spectrophotometerc was used to determine SUN, creatinine, phosphate, total calcium, total protein, albumin, globulin, cholesterol, and glucose concentrations and alkaline phosphatase, GGT, and alanine aminotranfersase activities. Serum sodium, potassium, and bicarbonate concentrations were determined by a selective-ion blood gas analyzer.d A spontaneously voided urine sample was collected into a sterile tube and underwent a urinalysis within 30 minutes after collection. A commercial dipsticke was dipped into the sample and visually read in accordance with the manufacturer's instructions. Urine specific gravity was determined by use of a manual veterinary optic refractometer.f The remaining urine sample was centrifuged at 1,372 × g for 3 minutes, and the supernatant was decanted. The urine sediment was resuspended in 0.3 to 0.5 mL of supernatant. A drop of the resulting suspension was placed on a microscope slide and microscopically evaluated. The presence of epithelial cells, casts, and mucus was evaluated at low (10×) power, and the presence of WBCs, RBCs, bacteria, fungi, crystals, parasites, sperm, and lipid was evaluated at high (40×) power. Urine protein and creatinine concentrations in the decanted supernatant were determined with an automated biochemical analyzerc and by use of a modified pyrogallol red-molybdate method and modified Jaffe procedure, respectively. γ-Glutamyltransferase activity, which is typically used to quantify GGT in the serum or plasma of humans, was used to spectrophotometricallyc determine urine GGT activity (expressed as U/L). Then, the UPC and urine GGT index (urine GGT activity [U/L]/urine creatinine concentration [mg/dL]; expressed as U/g) were calculated.
IOX administration and blood sample collection
For each rabbit, a 22-gauge catheter was aseptically placed into an auricular vein. Then, a commercially available IOX formulationg (64.7 mg/kg [0.1 mL/kg]) was administered as an IV bolus over 1 minute. The syringe and needle used to infuse the IOX were weighed before and after injection to determine the exact dose administered. A blood sample (0.5 mL) was collected by jugular venipuncture into a blood collection tube containing lithium heparin as an anticoagulant immediately before (0 minutes) and at 5, 15, 60, 90, 180, and 240 minutes after IOX administration. Each blood sample was centrifuged at 1,372 × g for 10 minutes. The plasma was harvested and stored in aliquots at −20°C until analysis.
HPLC method
The HPLC system consisted of a gradient pumph coupled to a UV detector,i which was set at 254 nm. The reverse-phase was conducted on a C18 columnj (particle size, 5 μm; length, 250 mm; internal diameter, 4.6 mm) that was maintained at room temperature (approx 22°C). A software programk was used for data processing.
The sample injection volume was 20 μL. The mobile phase consisted of 5% to 95% (vol/vol) acetonitrilel-water (pH, 2.7; acidified by the addition of 85% phosphoric acid in a dropwise manner). Both IOX and the ISg were eluted as 2 isomers. For analysis, the peak areas of the major IOX and IS isomers were used because they constituted >80% of the combined peak area, and the ratio of both isomer peaks remained constant at various IOX and IS concentrations under the described analytical conditions. For each sample, the peak area ratio was calculated as the ratio of the larger IOX peak to the larger IS peak by use of a software program,m and the peak area ratio was used for all subsequent calculations.
Preparation of plasma samples for HPLC—For each plasma sample in preparation for HPLC, 50 μL of the sample was added to 50 μL of a water-diluted IS solution (IS concentration, 50 μg/mL) and vigorously vortexed for 60 seconds. The mixture was then deproteinized by the addition of 100 μL of dichloromethane that had been diluted with 150 μL of double-distilled water, vigorously vortexed for 60 seconds, and centrifuged at 1,372 × g for 10 minutes. Twenty microliters of the resulting supernatant was decanted and centrifuged at 1,372 × g for 10 minutes before being injected into the HPLC system.
Stock solutions—Stock solutions of IOX and the IS were prepared monthly. Each stock solution had an IOX concentration of 1 mg/mL and was created by dilution of the parent compound in double-distilled water. The stock solutions were stored at −20°C until used. Iohexol working solutions were prepared fresh daily by further dilution of the IOX stock solutions in drug-free rabbit plasma. Seven working solutions with IOX concentrations of 2.5, 10, 25, 50, 125, 250, and 500 μg/mL were created each day and served as calibrators. Three in-house quality control solutions that contained IOX at low (25 μg/mL), medium (125 μg/mL), and high (500 μg/mL) concentrations were also prepared with rabbit plasma for assay validation. Aliquots of the IS stock solutions were diluted with water to produce working IS solutions with an IOX concentration of 50 μg/mL. Aliquots of the calibrators and quality control solutions were stored at −20°C until use. Nine standard curves were prepared, and all calibrators or quality control samples were injected in triplicate.
Validation—The HPLC method was validated in accordance with international guidelines.10,11 The specificity, sensitivity, linearity, LOD, LOQ, repeatability, and reproducibility were determined for the protocol. To determine linearity, standard solutions of IOX at concentrations ranging from 0.5 to 100 μg/mL (created by dilution of IOX stock solution with water) were analyzed to create calibration curves. Blank rabbit plasma samples spiked with IOX at concentrations of 2.5, 10, 25, 50, 125, 250, and 500 μg/mL were analyzed with the HPLC method. When dilution was accounted for, those spiked samples corresponded to IOX standard concentrations of 0.5, 2, 5, 10, 25, 50, and 100 μg/mL. The experiment was repeated 9 times. To evaluate specificity, blank rabbit plasma samples (ie, samples that contained no IOX or IS) were analyzed to check for the presence of interfering peaks at the elution times for IOX and IS isomers. Repeatability was assessed by the analysis of blank rabbit plasma samples spiked with IOX at concentrations of 25, 125, or 500 μg/mL, which corresponded to IOX standard concentrations of 5, 25, and 50 μg/mL, respectively. All samples were analyzed in triplicate on the same day. To determine the within-laboratory reproducibility, the repeatability experiment was repeated in triplicate on each of 7 days. Results of those experiments were also used to determine IOX recovery. The analytical recovery of IOX was assessed by comparison of the peak area ratio of spiked samples with the peak area ratio (analyte peak area/IS peak area) of the reference standards, which were prepared in water. The sensitivity of the HPLC method was expressed as the LOQ, which was defined as the minimum concentration of IOX in plasma that could be quantitatively determined with a peak height-to-baseline ratio of at least 10:1, and the LOD, which had a peak height-to-baseline ratio of 3:1.
To evaluate stability, aliquots of spiked samples underwent 3 freeze-thaw cycles. For each cycle, samples were frozen at −20°C for 24 hours and then allowed to thaw unassisted at room temperature. For the short-term stability test, IOX-spiked plasma aliquots (aliquots with known IOX concentrations of 25, 125, and 500 μg/mL) were thawed at room temperature and maintained at that temperature for 6 hours (ie, the duration of analysis for a typical batch) before analysis. For the long-term stability test, IOX-spiked plasma aliquots were thawed at room temperature and maintained at that temperature for 12 and 24 hours before analysis. Within each time frame, each IOX concentration was replicated in triplicate.
Pharmacokinetic and statistical analyses
The data distributions for CBC, serum biochemical, and urinalysis variables and the UPC and urine GGT index were assessed for normality by the Kolmogorov-Smirnov test.n All variables were normally distributed, and the results for each were summarized as the mean ± SD. Pharmacokinetic analyses were performed with commercial software.o Plasma IOX concentration data underwent noncompartmental analysis with a statistical moment approach. The AUC was calculated by the trapezoidal rule with extrapolation to infinity.12 The λ was calculated by extrapolation of the plasma concentration-time curve from the last measured concentration to infinity and use of at least 3 measured concentration points from the concentration-time curve that were selected by the software. Plasma Cl was calculated as the IOX dose administered/AUC; the IOX dose administered was estimated with the assumption that 85% of IOX injected was in the exo-IOX isoform. The volume of distribution was calculated as the plasma Cl/λ. The half-life was calculated from the λ of the terminal phase. The GFR was calculated as plasma Cl/subject body weight. Descriptive statistics were generated for each pharmacokinetic parameter.
The GFR estimated from the plasma Cl of IOX as determined from the plasma IOX concentration at 6 points (5, 15, 60, 90, 180, and 240 minutes after IOX administration) on the concentration-time curve was considered the referent against which the GFR estimated from each of 5 limited-sampling models was compared. For each limited-sampling model, the plasma IOX concentrations for various combinations of sampling times were used to calculate the Cl, which was subsequently used to calculate the GFR for that model. Five limited-sampling models (A, B, C, D, and E) were selected from all possible limited-sampling models for further analysis. Model A included the plasma IOX concentrations measured at 5, 15, 60, 90, and 180 minutes after IOX administration. Model B included plasma IOX concentrations measured at 5, 15, 60, and 90 minutes after IOX administration. Model C included plasma IOX concentrations measured at 5, 15, and 60 minutes after IOX administration. Model D included plasma IOX concentrations measured at 5, 15, and 90 minutes after IOX administration. Model E included plasma IOX concentrations measured at 5, 60, and 90 minutes after IOX administration. The correlation, association, and extent of agreement between the GFR as calculated for the referent model and the GFR calculated for each of the 5 limited-sampling models were assessed by calculation of the Pearson correlation coefficient, linear regression analysis, and Bland-Altman analysis, respectively.
Additionally, the accuracy of the GFR estimates for each limited-sampling model was determined as the percentage of study subjects for which the GFR calculated by that model did not deviate from the GFR estimated by the referent model by > 15%, 30%, and 50%. For each limited-sampling model and accuracy threshold (> 15%, 30%, and 50%), the percentage of subjects with GFR agreement between the given model and referent model was assessed with a χ2 test. Values of P < 0.05 were considered significant for all analyses.
Results
Rabbits
The study population consisted of 6 rabbits (2 sexually intact males, 1 sexually intact female, and 3 neutered females) with a mean ± SD age of 1.9 ± 1.6 years and body weight of 1.7 ± 0.2 kg. Descriptive statistics for CBC, serum biochemical, and urinalysis variables prior to IOX administration were summarized (Table 1). No adverse effects were observed in any rabbit during or after IOX administration.
Mean ± SD values for CBC, serum biochemical, and urinalysis variables for 6 healthy domestic rabbits (Oryctolagus cuniculus) enrolled in a study to evaluate GFR estimation by means of plasma Cl of IOX.
| Test | Variable | Mean ± SD | Reference range |
|---|---|---|---|
| CBC | RBC (× 1012 RBCs/L) | 5.5 ± 0.4 | 4–7 |
| Hct (%) | 34 ± 4.1 | 33–48 | |
| HGB (g/L) | 114 ± 111 | 100–150 | |
| MCV (fL) | 61.1 ± 2.7 | 60–75 | |
| MCH (fmol) | 1.3 ± 1.3 | 1.2–1.4 | |
| MCHC (mmol/L) | 2 ± 2.1 | 1.8–2.1 | |
| Platelet (× 109 platelets/L) | 292.5 ± 332 | 250–600 | |
| WBC (× 109 WBCs/L) | 5.4 ± 6.5 | 5–12 | |
| Serum biochemical profile | Creatinine (μmol/L) | 115 ± 8.8 | 44.2–221 |
| SUN (mmol/L) | 43.8 ± 8 | 28–63 | |
| Calcium (mmol/L) | 3 ± 0.1 | 1.4–3.1 | |
| Phosphorus (mmol/L) | 1.4 ± 0.2 | 1.3–2.2 | |
| Sodium (mmol/L) | 145,7 ± 4.3 | 131–155 | |
| Potassium (mmol/L) | 4.9 ± 0.3 | 3.6–6.9 | |
| Bicarbonate (mmol/L) | 26.8 ± 2.7 | 16–38 | |
| Total protein (g/dL) | 66 ± 3 | 54–83 | |
| Albumin (g/L) | 48 ± 3 | 24–46 | |
| Globulin (g/L) | 18 ± 3 | 15–28 | |
| ALKP (U/L) | 166 ± 91 | 66–266 | |
| GGT (U/L) | 108 ± 38 | 0–116 | |
| ALT (U/L) | 866 ± 260 | 800–1,333 | |
| Cholesterol (mmol/L) | 0.1 ± 0.2 | 0.2–2 | |
| Glucose (mmol/L) | 8.5 ± 1.7 | 4.1–8.6 | |
| Urinalysis | USG | 1.034 ± 0.019 | 1.003–1.036 |
| pH | 8.3 ± 0.5 | 8.2 | |
| UPC | 0.3 ± 0.2 | 0.1–0.4 | |
| GGT index (U/g) | 0.81 ± 0.49 | 0.04–1.03 |
Blood and urine samples were obtained before IOX administration.
ALKP = Alkaline phosphatase. ALT = Alanine aminotranfersase. HGB = Hemoglobin. MCH = Mean corpuscular hemoglobin. MCHC = Mean corpuscular hemoglobin concentration. MCV = Mean corpuscular volume. USG = Urine specific gravity.
Validation of HPLC method
Iohexol was eluted as 2 isomers, endo-IOX and exo-IOX, which had elution times of 6.4 and 6.8 minutes, respectively. The IS was also eluted as 2 isomers, which had elution times of 10.4 and 11.0 minutes (Figure 1). Analysis of rabbit plasma samples collected prior to IOX administration revealed that there were no interfering peaks at the elution times for the IOX and IS isomers, which indicated that the HPLC method was specific for identification of IOX and the IS in rabbit plasma samples. The mean ± SD percentage recovery for the IS ranged from 94.2 ± 3.1% to 96.5 ± 2.0%. The mean ± SD coefficient of determination for the HPLC method was 0.997 ± 0.006, which indicated the assay had almost perfect linearity across the IOX concentration range of the calibrator samples. The LOD was 0.01 μg/mL, and the LOQ was 0.10 μg/mL. Results of the repeatability, reproducibility, and recovery experiments to validate the HPLC method were summarized (Table 2). The measured IOX concentration in spiked plasma samples following 3 freeze-thaw cycles did not differ significantly from the starting concentration (ie, the percentage recovery did not differ significantly from 100%; Table 3), which indicated that IOX was fairly stable in rabbit plasma and was not affected by freezing and thawing.
Representative chromatogram for a working solution of blank rabbit plasma that was spiked with a commercial IOX formulation to achieve an IOX concentration of 10 μg/mL. The IS was the same commercial IOX formulation used to spike the blank plasma sample that was diluted with water to achieve an IOX concentration of 50 μg/mL. The IOX in the spiked rabbit plasma sample was eluted as 2 isomers, endo-IOX and exo-IOX, at 6.4 and 6.8 minutes, respectively, after injection of the sample into the HPLC unit. The IOX in the IS was also eluted as 2 isomers at 10.4 and 11.0 minutes after injection into the HPLC unit. Analysis of blank rabbit plasma samples that were not spiked with IOX revealed that there were no interfering peaks at the elution times for the IOX and IS isomers. mAU = Milli-arbitrary units.
Citation: American Journal of Veterinary Research 80, 6; 10.2460/ajvr.80.6.525
Results of repeatability and reproducibility experiments conducted to validate the HPLC method used to measure IOX concentration in rabbit plasma in a study to evaluate GFR estimation by means of plasma Cl of IOX.
| Assay measure | Known IOX concentration (μg/mL) | Mean ± SD measured IOX concentration (μg/mL) | Coefficient of variation (%) | Mean ± SD recovery (%) |
|---|---|---|---|---|
| Repeatability | 25 | 24.92 ± 0.08 | 1.33 | — |
| 125 | 124.10 ± 0.22 | 2.80 | — | |
| 500 | 498.12 ± 3.86 | 7.12 | — | |
| Reproducibility | 25 | 24.57 ± 0.67 | 2.45 | 96.01 ± 2.50 |
| 125 | 124.51 ± 0.67 | 5.51 | 95.32 ± 1.37 | |
| 500 | 488.23 ± 1.55 | 2.00 | 95.04 ± 3.13 |
For the repeatability experiment, each quality control IOX solution (ie, solution with known IOX concentration) was measured in triplicate on 1 day; therefore, each measured value represents the mean ± SD for 3 replicates. For the reproducibility experiment, the repeatability experiment was repeated on each of 7 days; therefore, each measured value represents the mean ± SD for 21 replicates. For each replicate of the reproducibility experiment, recovery was calculated as the measured IOX concentration/known IOX concentration; therefore, each presented value represents the mean ± SD for 21 replicates.
— = Not calculated.
Mean ± SD percentage of IOX recovery from rabbit plasma samples that were spiked with a known concentration of the compound after 3 freeze-thaw cycles and maintenance of samples at room temperature (approx 22°C) for 6, 12, and 24 hours before HPLC analysis.
| No. of hours that thawed samples were maintained at room temperature before analysis | |||
|---|---|---|---|
| Known IOX concentration (μg/mL) | 6 | 12 | 24 |
| 25 | 97.00 ± 1.00 | 99.50 ± 1.29 | 99.76 ± 2.10 |
| 125 | 96.33 ± 1.15 | 101.01 ± 0.82 | 97.00 ± 1.89 |
| 500 | 96.67 ± 2.08 | 97.03 ± 2.50 | 98.01 ± 2.35 |
Each value represents the mean ± SD percentage for 3 replicates. For each freeze-thaw cycle, samples were frozen at −20°C for 24 hours and then allowed to thaw unassisted at room temperature. For the short-term stability test, IOX-spiked sample aliquots were thawed at room temperature and maintained at that temperature for 6 hours (ie, the duration of analysis for a typical batch) before analysis. For the long-term stability tests, IOX-spiked sample aliquots were thawed at room temperature and maintained at that temperature for 12 and 24 hours before analysis.
IOX pharmacokinetics and estimated GFR
The mean ± SD plasma IOX concentration over time was plotted (Figure 2). The IOX pharmacokinetic parameters and estimated GFR for the study rabbits were summarized (Table 4).
Mean ± SD plasma IOX concentration over time for 6 healthy conscious and unsedated domestic rabbits (Oryctolagus cuniculus) following administration of a commercial IOX formulation (64.7 mg/kg [0.1 mL/kg], IV as a slow bolus over 1 minute).
Citation: American Journal of Veterinary Research 80, 6; 10.2460/ajvr.80.6.525
Pharmacokinetic parameters for IOX following administration of a single dose (64.7 mg/kg [0.1 mL/kg], IV as a slow bolus over 1 minute) to each of 6 healthy domestic rabbits and the estimated GFR for those rabbits.
| Parameter | Mean ± SD |
|---|---|
| Total IOX dose administered (mg/kg) | 108 ± 15 |
| t1/2 (min) | 47 ± 11 |
| Cl (mg/[μg/mL]/min) | 0.007 ± 0.001 |
| AUC (μg/mL·min) | 15,451 ± 3,716 |
| Vd (mg/μg/mL) | 0.45 ± 0.20 |
| GFR (mL/min/kg) | 4.41 ± 1.10 |
t1/2 = Half-life. Vd = Volume of distribution.
The Pearson correlation coefficient (r), coefficient of determination (R2), and linear regression slope and intercept for comparison of the GFR as estimated by each of 5 limited-sampling models relative to the GFR estimated by the 6-point referent model were summarized (Table 5). There was a strong positive correlation between the GFR estimated by each of the limited-sampling models and that estimated by the referent model. Also, the GFR estimated by each of the simplified models did not differ significantly from the GFR estimated by the referent model.
Pearson correlation coefficient (r), coefficient of determination (R2), and linear regression slope and intercept for the GFR estimated by each of 5 limited-sampling models relative to the GFR estimated by a 6-point model (referent).
| Model | |||||
|---|---|---|---|---|---|
| Variable | A | B | C | D | |
| r | 0.9976 | 0.9812 | 0.9378 | 0.9640 | 0.9264 |
| R2 | 0.9952 | 0.9628 | 0.8794 | 0.9293 | 0.8583 |
| Slope | 1.004 | 1.153 | 1.215 | 0.9679 | 0.8844 |
| Intercept | 0.00008 | −0.00042 | −0.00065 | 0.00028 | 0.00068 |
The GFR estimated from the plasma Cl of IOX as determined from the plasma IOX concentration at 6 points (5, 15, 60, 90, 180, and 240 minutes after IOX administration) on the concentration-time curve was considered the referent against which the GFR estimated for each of the limited-sampling models was compared. For each limited-sampling model, the plasma IOX concentrations for various combinations of sampling times were used to calculate the Cl, which was subsequently used to calculate the GFR for that model. Model A included the plasma IOX concentrations measured at 5, 15, 60, 90, and 180 minutes after IOX administration. Model B included plasma IOX concentrations measured at 5, 15, 60, and 90 minutes after IOX administration. Model C included plasma IOX concentrations measured at 5, 15, and 60 minutes after IOX administration. Model D included plasma IOX concentrations measured at 5, 15, and 90 minutes after IOX administration. Model E included plasma IOX concentrations measured at 5, 60, and 90 minutes after IOX administration.
The mean ± SD bias and 95% limits of agreement between the GFR estimated by each of the simplified models and the GFR estimated by the referent model were summarized (Table 6). Model D had the smallest magnitude of mean bias and was selected as the preferred limited-sampling method for estimating GFR on the basis of plasma IOX Cl. For model D, the GFR was estimated from the plasma IOX concentrations at 5, 15, and 90 minutes after IOX administration; thus, it was a 3-point model. The mean ± SD GFR as estimated by model D was 4.46 ± 1.26 mL/min/kg (range, 3.39 to 7.02 mL/min/kg).
Results of Bland-Altman analyses in which the GFR estimated by each of the 5 limited-sampling models described in Table 5 was compared with the GFR estimated by the referent model.
| Accuracy threshold* | |||||
|---|---|---|---|---|---|
| Model | Mean ± SD bias in GFR relative to referent (mL/min kg) | 95% limits of agreement for the bias in GFR (mL/min kg) | 15% | 30% | 50% |
| A | −0.08 ± 0.06 | −0.19 to 0.04 | 100 | 98 | 98 |
| B | −0.42 ± 0.22 | −0.86 to 0.02 | 100 | 95 | 96 |
| C | −0.55 ± 0.33 | −1.20 to 0.11 | 100 | 100 | 98 |
| D | −0.05 ± 0.19 | −0.42 to 0.32 | 100 | 95 | 96 |
| E | 0.08 ± 0.24 | −0.38 to 0.55 | 100 | 95 | 96 |
Percentage of study subjects for which the GFR estimated by the given model did not deviate from the GFR estimated by the referent model by greater than the given threshold. Within each threshold, the accuracy did not differ significantly among the limited-sampling models.
See Table 5 for remainder of key.
Discussion
Glomerular filtration rate is universally considered the best indicator of overall renal function in both human13 and veterinary2,4,7,9,14,15 patients. Plasma Cl of IOX is a reliable and fairly easy method to estimate GFR in various veterinary species.2,6,7 Although use of the plasma Cl of IOX is becoming a popular technique for estimation of GFR and the early diagnosis of CKD in dogs and cats, its use had not been described in domestic rabbits prior to the present study. In experimental settings, the GFR of rabbits has been estimated by plasma Cl of iodinaxol3 and inulin.16 In the present study, healthy unsedated domestic rabbits received a single injection of IOX (64.7 mg/kg, IV), and the plasma Cl of IOX was used to estimate GFR. The dose of IOX administered was well tolerated by the rabbits, and no adverse effects were observed in any rabbit during or after IOX administration. The HPLC method used to measure plasma IOX concentration in the present study proved to be reliable and accurate and should be applicable in routine clinical settings. However, the reproducibility of the method was assessed by use of a very small data set, which was not consistent with guidelines established by the American Society for Veterinary Clinical Pathology.11 Nevertheless, IOX is a safe, fairly inexpensive, and readily available tracer agent.
Traditionally, measurement of serum creatinine and SUN concentrations and calculation of the UPC are used to assess renal function in rabbits.17 Those 3 variables can be substantially affected by several extrarenal factors. In rabbits, SUN has a circadian rhythm and is highly dependent on the protein intake and nutritional status of the animal. Relative to other mammals, rabbits have a reduced capacity to concentrate urea, and intestinal absorption, cecal flora activity, liver function, gastrointestinal tract hemorrhage, stress, and hydration status can affect SUN concentration.18 Creatinine is a product of protein catabolism that is freely filtered through renal glomeruli and is excreted at a constant rate. Creatinine is considered a more reliable indicator of renal function than SUN because it is less likely to be affected by extrarenal factors.17 However, creatinine is an insensitive indicator of modest renal damage because it increases from the reference range only when > 50% to 70% of nephrons become impaired and cease to function.17 Information regarding the use of exogenous creatinine Cl for estimation of renal function is lacking for rabbits.
In the present study, the mean ± SD GFR was 4.41 ± 1.10 mL/min/kg when estimated from the plasma IOX concentrations obtained at 5, 15, 60, 90, 180, and 240 minutes after IOX administration (ie, 6-point model [referent]). That GFR estimate was consistent with the GFR estimated for healthy rabbits by plasma Cl of inulin (4.01 ± 0.14 mL/min/kg)16 and for healthy male rabbits by plasma clearance of iodixanol (by means of a 3-point model; 4.21 ± 0.28 mL/min/kg)3 in other studies. To our knowledge, the present study was the first to describe estimation of GFR on the basis of plasma Cl of IOX. Iohexol is a nonionic, low-osmolarity contrast medium that is widely used in many species because of its low toxicity.2 Use of plasma Cl of IOX for estimation of GFR has a practical advantage over use of plasma Cl of inulin because IOX can be administered as a single IV bolus, whereas inulin requires continuous infusion and is therefore more stressful for the patient or study subject.
In the present study, 3 different 3-point limited-sampling models were evaluated on the basis of results of other studies2,4,14 in which plasma Cl of IOX was used to estimate GFR in veterinary species other than rabbits. Although the GFR estimated by each of the 3-point limited-sampling models did not differ significantly from the GFR estimated by the 6-point referent model, we considered model D, which was based on the plasma IOX concentrations at 5, 15, and 90 minutes after IOX administration, to be the best alternative in terms of agreement with the referent model and clinical feasibility (requiring the collection of only 3 blood samples within 90 minutes after IOX administration). The mean ± SD estimated GFR for model D (4.46 ± 1.26 mL/min/kg) was also similar to the GFR estimated for healthy rabbits in the previously described studies.3,16
Results of the present study indicated that, in healthy rabbits, the GFR could be easily and accurately estimated by measurement of the plasma IOX concentration at 5, 15, and 90 minutes after injection of the tracer (ie, a 3-point plasma IOX Cl model). Use of plasma Cl of IOX for estimation of GFR does not require continuous IV infusion of the tracer or collection of urine, which is advantageous in clinical settings. Additionally, no adverse effects were associated with IOX administration in the rabbits of the present study. Therefore, plasma Cl of IOX may be clinically beneficial for estimation of GFR and early detection of CKD in rabbits. However, further research is necessary to evaluate whether GFR of rabbits is affected by breed, sex, or age and to determine whether plasma Cl of IOX in rabbits varies at various stages of CKD.
ABBREVIATIONS
| X | Terminal slope of the concentration-time curve |
| AUC | Area under the concentration-time curve |
| CKD | Chronic kidney disease |
| Cl | Clearance |
| GFR | Glomerular filtration rate |
| GGT | γ-Glutamyltransferase |
| HPLC | High-performance liquid chromatography |
| IOX | Iohexol |
| IS | Internal standard |
| LOD | Limit of determination |
| LOQ | Limit of quantification |
| UPC | Urine protein-to-creatinine ratio |
Footnotes
ELISA, Medicago, Uppsala, Sweden.
Procyte DX, Idexx, Milano, Italy.
SLIM, Seac, Firenze, Italy.
STAT Profile pHOx Plus, GEPA, Milano, Italy.
VetLab UA, Idexx Europe BV, Hoofddorp, Netherlands.
Vet 360 refractometer, Reichert Technologies, Buffalo, NY.
Omnipaque 350, Nycomed Imaging AS, Oslo, Norway.
Pump Series 200, PerkinElmer Inc, Nashville, Tenn.
Series 200 variable UV detector, PerkinElmer Inc, Nashville, Tenn.
HAISIL HL, The Nest Group Inc, Southborough, Mass.
Turbochrom, PerkinElmer Inc, Nashville, Tenn.
VWR International, Milan, Italy.
Office 2007, Microsoft Corp, Redmond, Wash.
Prism, version 5.0a, GraphPad Software Inc, La Jolla, Calif.
Zhang Y, Huo M, Zhou J, et al. PKSolver: an add-in program for pharmacokinetic and pharmacodynamic data analysis in Microsoft Excel. Comput Methods Programs Biomed 2010;99:306–314.
References
1. Harcourt-Brown FM. Diagnosis of renal disease in rabbits. Vet Clin North Am Exot Anim Pract 2013;16:145–174.
2. Meucci V, Sgorbini M, Bonelli F, et al. Determination of glomerular filtration rate in adult horses and donkeys by single IV administration of iohexol. J Equine Vet Sci 2015;35:36–40.
3. Michigoshi Y, Katayama R, Yamagishi N, et al. Estimation of glomerular filtration rate in rabbits by a single-sample method using iodixanol. Lab Anim 2012;46:341–344.
4. Goy-Thollot I, Chafotte C, Besse S, et al. Iohexol plasma clearance in healthy dogs and cats. Vet Radiol Ultrasound 2006;47:168–173.
5. Lefebvre H. Renal function testing. In: Bartges J, Polzin D, eds. Nephrology and urology of small animals. West Sussex, England: Wiley-Blackwell UK, 2011;91–96.
6. Lippi I, Meucci V, Guidi G, et al. Glomerular filtration rate evaluation in the dog throughout the plasmatic clearance of iohexol: simplified methods. Veterinaria 2008;22:53–60.
7. Meucci V, Guidi G, Melanie P, et al. A limited sampling, simple, and useful method for determination of glomerular filtration rate in cats by using a new accurate HPLC method to measure iohexol plasmatic concentrations. J Vet Med 2013;2013:569121.
8. Brown SC, 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.
9. 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.
10. European Medicines Agency. Guideline on bioanalytical method validation. 2011. Available at: www.ema.europa.eu/documents/scientific-guideline/guideline-bioanalytical-method-validation_en.pdf. Accessed Nov 29, 2018.
11. Flatland B, Freeman KP, Friedrichs KR, et al. ASVCP quality assurance guidelines: control of general analytical factors in veterinary laboratories. Vet Clin Pathol 2010;39:264–277.
12. Gibaldi M, Perrier D. Pharmacokinetics. 2nd ed. New York: Marcel Dekker, 1982;445–449.
13. Schwartz GJ, Furth SL. Glomerular filtration rate measurement and estimation in chronic kidney disease. Pediatr Nephrol 2007;22:1839–1848.
14. Meucci V, Gasperini A, Soldani G, et al. A new HPLC method to determine glomerular filtration rate and effective renal plasma flow in conscious dogs by single intravenous administration of iohexol and p-aminohippuric acid. J Chromatogr Sci 2004;42:107–111.
15. Bexfield NH, Heiene R, Gerritsen RJ, et al. Glomerular filtration rate estimated by 3-sample plasma clearance of iohexol in 118 healthy dogs. J Vet Intern Med 2008;22:66–73.
16. Michigoshi Y, Yamagishi N, Satoh H, et al. Using a single blood sample and inulin to estimate glomerular filtration rate in rabbits. J Am Assoc Lab Anim Sci 2011;50:702–707.
17. Mancinelli E, Shaw DJ, Meredith AL. γ-Glutamyl-transferase (GGT) activity in the urine of clinically healthy domestic rabbits (Oryctolagus cuniculus). Vet Rec 2012;171:475.
18. Jenkins JR. Rabbit diagnostic testing. J Exot Pet Med 2008;17:4–15.
