• View in gallery
    Figure 1—

    Scatterplot (A) and Bland-Altman plot (B) of global GFR assessment determined by use of PC-GFR and CT-GFR in 8 cats. Each square represents results for 1 cat. In the scatterplot, the diagonal line is the line of perfect agreement. In the Bland-Altman plot, the solid horizontal line indicates the mean difference between the methods and the dashed horizontal lines indicate the 95% limits of agreement (ie, ± 1.96 SD).

  • View in gallery
    Figure 2—

    Scatterplots (A and B) and Bland-Altman plots (C and D) of the percentage GFR for the left (A and C) and right (B and D) kidney determined by use of NM-L-GFR and CT-L-GFR in 8 cats. The GFR for each kidney is reported as the percentage of global GFR. See Figure 1 for remainder of key.

  • 1.

    Polzin DJ, Osborne CA & Ross S. Chronic kidney disease. In: Ettinger SJ, Feldman EC, eds. Textbook of veterinary internal medicine. Vol II. 6th ed. St Louis: Elsevier Saunders, 2005;17561785.

    • Search Google Scholar
    • Export Citation
  • 2.

    DiBartola SP. Renal disease: clinical approach and laboratory evaluation. In: Ettinger SJ, Feldman EC, eds. Textbook of veterinary internal medicine. Vol II. 6th ed. St Louis: Elsevier Saunders, 2005;17161730.

    • Search Google Scholar
    • Export Citation
  • 3.

    Muller-Surr R. Radiopharmaceuticals: their intrarenal handling and localization. In: Murray PC, Ell PJ, eds. Nuclear medicine in clinical diagnosis and treatment. Vol I. New York: Churchill Livingston, 2004;195212.

    • Search Google Scholar
    • Export Citation
  • 4.

    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; 65:15021507.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5.

    Bailey DB, Rassnick KM, Dykes NL, et al. Phase I evaluation of carboplatin by use of a dosing strategy based on a targeted area under the platinum concentration-versus-time curve and individual glomerular filtration rate in cats with tumors. Am J Vet Res 2009; 70:770776.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6.

    Bailey DB, Rassnick KM, Prey JD, et al. Evaluation of serum iohexol clearance for use in predicting carboplatin clearance in cats. Am J Vet Res 2009; 70:11351140.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 7.

    Barthez PY, Hornof WJ, Cowgill LD, et al. Comparison between the scintigraphic uptake and plasma clearance of 99mTc-diethylenetriaminepentacetic acid (DTPA) for the evaluation of the glomerular filtration rate in dogs. Vet Radiol Ultrasound 1998; 39:470474.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8.

    Kyles AE, Hardie EM, Wooden BG, et al. Clinical, clinicopathologic, radiographic, and ultrasonographic abnormalities in cats with ureteral calculi: 163 cases (1984–2002). J Am Vet Med Assoc 2005; 226:932936.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9.

    Hackstein N, Cengiz H & Rau W. Contrast media clearance in a single kidney measured on multiphasic helical CT: results in 50 patients without acute renal disorder. AJR Am J Roentgenol 2002; 178:111118.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10.

    Hackstein N, Bauer J, Hauck EW, et al. Measuring single-kidney glomerular filtration rate on single-detector helical CT using a two-point Patlak plot technique in patients with increased interstitial space. AJR Am J Roentgenol 2003; 181:147156.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11.

    Hackstein N, Cengiz H, Rau W, et al. Glomerular filtration rate measured by using triphasic helical CT with a two-point Patlak plot technique. Radiology 2004; 230:221226.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12.

    Tsushima Y, Blomley MJ, Okabe K, et al. Determination of glomerular filtration rate per unit renal volume using computerized tomography: correlation with conventional measures of total and divided renal function. J Urol 2001; 165:382385.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13.

    O'Dell-Anderson KJ, Twardock R, Grimm JB, et al. Determination of glomerular filtration rate in dogs using contrast-enhanced computed tomography. Vet Radiol Ultrasound 2006; 47:127135.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14.

    Alexander K, Del Castillo JR, Ybarra N, et al. Single-slice dynamic computed tomographic determination of glomerular filtration rate by use of Patlak plot analysis in anesthetized pigs. Am J Vet Res 2007; 68:297304.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15.

    Patlak CS, Blasberg RG, Fenstermacher JD. Graphical evaluation of blood-to-brain transfer constants from multiple-time uptake data. J Cereb Blood Flow Metab 1983; 3:17.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16.

    Twardock AR & Bahr A. Renal scintigraphy. In: Daniel GB, Berry CR, eds. Textbook of veterinary nuclear medicine. 2nd ed. Knoxville, Tenn: American College of Veterinary Radiology, 2006;332333.

    • Search Google Scholar
    • Export Citation
  • 17.

    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; 61:280285.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18.

    Altman DG, Bland JM. Measurement in medicine: the analysis of method comparison studies. Statistician 1983; 32:307317.

  • 19.

    Dewitte K, Fierens C, Stöckl D, et al. Application of the Bland-Altman plot for interpretation of method-comparison studies: a critical investigation of its practice. Clin Chem 2002; 48:799801.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20.

    Alexander K, Dunn M, Carmel EN, et al. Clinical application of Patlak plot CT-GFR in animals with upper urinary tract disease. Vet Radiol Ultrasound 2010; 51:421427.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21.

    Altman DG. Some common problems in medical research. In: Practical statistics for medical research. London: Chapman & Hall/CRC, 1997;396439.

    • Search Google Scholar
    • Export Citation
  • 22.

    Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 327:307310.

  • 23.

    Fusellier M, Desfontis J, Madec S, et al. Influence of three anesthetic protocols on glomerular filtration rate in dogs. Am J Vet Res 2007; 68:807811.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24.

    Newell SM, Ko JC, Ginn PE, et al. Effects of three sedative protocols on glomerular filtration rate in clinically normal dogs. Am J Vet Res 1997; 58:446450.

    • Search Google Scholar
    • Export Citation
  • 25.

    Morcos SK. Contrast medium-induced nephrotoxicity. In: Dawson P, Cosgrove DO, Grainger RG, eds. Textbook of contrast media. Oxford, England: Isis Medical Media, 1999;135148.

    • Search Google Scholar
    • Export Citation
  • 26.

    Daghini E, Juillard L, Haas JA, et al. Comparison of mathematic models for assessment of glomerular filtration rate with electron-beam CT in pigs. Radiology 2007; 242:417424.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 27.

    Chen AY, Zirwas MJ, Heffernan MP. Nephrogenic systemic fibrosis: a review. J Drugs Dermatol 2010; 9:829834.

  • 28.

    Mendichovszky I, Pedersen M, Frøkiær J, et al. How accurate is dynamic contrast-enhanced MRI in the assessment of renal glomerular filtration rate? A critical appraisal. J Magn Reson Imaging 2008; 27:925931.

    • Crossref
    • Search Google Scholar
    • Export Citation

Advertisement

Comparison of glomerular filtration rate determined by use of single-slice dynamic computed tomography and scintigraphy in cats

David M. SchmidtDepartment of Clinical Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY 14853.

Search for other papers by David M. Schmidt in
Current site
Google Scholar
PubMed
Close
 DVM
,
Peter V. ScrivaniDepartment of Clinical Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY 14853.

Search for other papers by Peter V. Scrivani in
Current site
Google Scholar
PubMed
Close
 DVM
,
Nathan L. DykesDepartment of Clinical Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY 14853.

Search for other papers by Nathan L. Dykes in
Current site
Google Scholar
PubMed
Close
 DVM
,
Richard M. GoldsteinDepartment of Clinical Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY 14853.

Search for other papers by Richard M. Goldstein in
Current site
Google Scholar
PubMed
Close
 DVM
,
Hollis N. ErbDepartment of Population Medicine and Diagnostic Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY 14853.

Search for other papers by Hollis N. Erb in
Current site
Google Scholar
PubMed
Close
 DVM, PhD
, and
Anthony P. ReevesSchool of Electrical and Computer Engineering, College of Engineering, Cornell University, Ithaca, NY 14853.

Search for other papers by Anthony P. Reeves in
Current site
Google Scholar
PubMed
Close
 PhD

Abstract

Objective—To compare estimation of glomerular filtration rate determined via conventional methods (ie, scintigraphy and plasma clearance of technetium Tc 99m pentetate) and dynamic single-slice computed tomography (CT).

Animals—8 healthy adult cats.

Procedures—Scintigraphy, plasma clearance testing, and dynamic CT were performed on each cat on the same day; order of examinations was randomized. Separate observers performed GFR calculations for scintigraphy, plasma clearance testing, or dynamic CT. Methods were compared via Bland-Altman plots and considered interchangeable and acceptable when the 95% limits of agreement (mean difference between methods ± 1.96 SD of the differences) were ≤ 0.7 mL/min/kg.

Results—Global GFR differed < 0.7 mL/min/kg in 5 of 8 cats when comparing plasma clearance testing and dynamic CT; the limits of agreement were 1.4 and −1.7 mL/min/kg. The mean ± SD difference was −0.2 ± 0.8 mL/min/kg, and the maximum difference was 1.6 mL/min/kg. The mean ± SD difference (absolute value) for percentage filtration by individual kidneys was 2.4 ± 10.5% when comparing scintigraphy and dynamic CT; the maximum difference was 20%, and the limits of agreement were 18% and 23% (absolute value).

Conclusions and Clinical Relevance—GFR estimation via dynamic CT exceeded the definition for acceptable clinical use, compared with results for conventional methods, which was likely attributable to sample size and preventable technical complications. Because 5 of 8 cats had comparable values between methods, further investigation of dynamic CT in a larger sample population with a wide range of GFR values should be performed.

Abstract

Objective—To compare estimation of glomerular filtration rate determined via conventional methods (ie, scintigraphy and plasma clearance of technetium Tc 99m pentetate) and dynamic single-slice computed tomography (CT).

Animals—8 healthy adult cats.

Procedures—Scintigraphy, plasma clearance testing, and dynamic CT were performed on each cat on the same day; order of examinations was randomized. Separate observers performed GFR calculations for scintigraphy, plasma clearance testing, or dynamic CT. Methods were compared via Bland-Altman plots and considered interchangeable and acceptable when the 95% limits of agreement (mean difference between methods ± 1.96 SD of the differences) were ≤ 0.7 mL/min/kg.

Results—Global GFR differed < 0.7 mL/min/kg in 5 of 8 cats when comparing plasma clearance testing and dynamic CT; the limits of agreement were 1.4 and −1.7 mL/min/kg. The mean ± SD difference was −0.2 ± 0.8 mL/min/kg, and the maximum difference was 1.6 mL/min/kg. The mean ± SD difference (absolute value) for percentage filtration by individual kidneys was 2.4 ± 10.5% when comparing scintigraphy and dynamic CT; the maximum difference was 20%, and the limits of agreement were 18% and 23% (absolute value).

Conclusions and Clinical Relevance—GFR estimation via dynamic CT exceeded the definition for acceptable clinical use, compared with results for conventional methods, which was likely attributable to sample size and preventable technical complications. Because 5 of 8 cats had comparable values between methods, further investigation of dynamic CT in a larger sample population with a wide range of GFR values should be performed.

Renal disease is a common disorder of older cats, and early identification of renal disease is important to prevent or slow progression to overt renal failure.1 Additionally, it is important to quantify renal function when making certain decisions regarding patient management (eg, whether to perform nephrectomy). Common examinations of renal function include assessment of serum BUN and creatinine concentrations; however, elevated concentrations are insensitive assessments for early renal disease because they only become indicative of disease when > 75% of total renal function is lost.2 An ideal test would be sensitive to smaller reductions in renal function, readily available, able to assess relative function in each kidney, and yield timely results. Currently available tests have some but not all of these features.

Evaluation of GFR is considered the best way to estimate renal function and is an important measure of early or mild renal disease.1 Many methods for estimation of GFR exist, and they all have advantages and disadvantages. The criterion-referenced method for estimation of GFR is considered to be assessment of urinary clearance of inulin3; however, this method is expensive and time-consuming, requires urinary catheterization, and is rarely used in a clinical setting. Furthermore, inulin clearance can only determine global GFR (rather than GFR of each kidney). Therefore, veterinarians typically rely on other methods to estimate GFR, such as determination of plasma clearance of creatinine or iohexol. Iohexol has properties similar to those of inulin in that iohexol is freely filtered by the glomeruli and not secreted or reabsorbed in the renal tubules. Iohexol is readily available and easily administered but also only assesses global GFR. Additionally, the iohexol in plasma samples must be assayed with high-performance liquid chromatography, and test results are currently available within 1 week. This is acceptable when managing chronic disease but not when immediate results could affect patient care (eg, unilateral nephrectomy, renal transplantation, and adjustments in chemotherapeutic dose).4–6

At our veterinary medical teaching hospital, estimation of GFR in cats is commonly performed by use of plasma clearance of 99mTc-DTPA and renal scintigraphy (to determine the percentage filtration of each kidney). Together, these methods yield estimations for both global GFR and GFR of each kidney. These tests are affordable, and results are available the same day. Another study7 revealed excellent agreement between PC-GFR and urinary clearance of inulin. However, these methods require use of radioisotopes; therefore, patients must be hospitalized to allow time to excrete the radiopharmaceutical prior to discharge. Additionally, these methods provide minimal morphological information and require the use of a gamma camera, which is of limited availability.

Estimation of GFR by use of dynamic CT has many advantages. In contrast to scintigraphy, dynamic CT does not require the use of injected radioisotopes. Thus, the procedure is safer for personnel and patients do not have to be hospitalized during radiopharmaceutical excretion. Furthermore, dynamic CT can be used to calculate GFR for each kidney, which is not possible with plasma clearance methods (eg, inulin and iohexol). Additionally, dynamic CT provides important morphological information about the kidneys and ureters (eg, signs of ureteral obstruction or renal neoplasm) that is not obtained by use of any of the global methods (ie, clearance of inulin, iohexol, or DTPA). In other words, dynamic CT can be used to quantify renal function and also may disclose the underlying cause of renal dysfunction. This advantage of dynamic CT may prove important because the incidence of ureteral obstruction in cats is thought to have increased in recent years, as was reported8 in at least 1 population of cats during the period between 1984 and 2002.

Estimation of GFR by use of CT is a relatively new method that has yielded promising results in humans.9–12 In dogs, GFR has been analyzed by dynamic single-slice CT, renal scintigraphy, and iohexol clearance, and CT has provided promising results for clinical use.13 In pigs, a comparison of CT-GFR and inulin clearance found no significant difference between methods for evaluation of global GFR and GFR of the right kidney.14 To our knowledge, studies to compare methods of assessing GFR have not been performed in cats, but cats are commonly affected by renal disease. Therefore, the purpose of the study reported here was to compare results for conventional methods used in our teaching hospital for the estimation of GFR (ie, PC-GFR, NM-L-GFR, and NM-R-GFR) with results for dynamic single-slice CT (ie, CT-GFR, CT-L-GFR, and CT-R-GFR) in cats.

Materials and Methods

Animals—Eight healthy adult cats were used in the study; data collection was conducted between April and May 2010. Privately owned (veterinary students and university personnel) cats were volunteered for use in the study. The sample population included 6 neutered males and 2 spayed females (7 domestic shorthair cats and 1 domestic longhair cat). Mean ± SD age was 3.7 ± 2.7 years (range, 1.0 to 8.4 years), and mean body weight was 5.8 ± 1.5 kg (range, 4.0 to 7.8 kg). Cats were deemed healthy on the basis of results of a physical examination. Consent was obtained from owners for participation of their cats. The study was approved by the Institutional Animal Care and Use Committee at Cornell University.

Cats were sedated with ketamine (5 to 10 mg/kg, IM) and midazolam (0.1 to 0.2 mg/kg, IM) to enable catheter placement. All cats had catheters inserted in a cephalic vein (for injection of iohexol for CT and radiopharmaceutical for scintigraphy) and in a lateral saphenous or external jugular vein (for collection of blood samples and administration of anesthetic). Blood samples were collected twice, and PCV was determined on each sample; the mean value for the 2 measurements was used for subsequent calculations.

Cats were anesthetized, and CT and scintigraphy were performed during the same anesthetic episode. Anesthesia consisted of a bolus injection of propofol (5 mg/kg, IV) followed by a continuous rate infusion (0.05 to 0.1 mg/kg/min, IV) for the duration of both procedures (approx 10 to 15 min/procedure). Computed tomography and scintigraphy were performed sequentially, on the same day. The order of examination was randomized by coin toss; thus, CT was performed first on 4 cats, and scintigraphy was performed first on the other 4 cats.

CT—All images were obtained by use of the same 16-slice helical CT scanner.a Image acquisition involved 3 steps. First, the cranial portion of the abdomen was scanned to identify an appropriate slice that included both kidneys and the aorta (the entire urinary tract was not evaluated to minimize patient exposure). Images were obtained by use of helical acquisition with 120 kVp, 50 mA, slice thickness of 1 mm, pitch of 1, and 512 × 512 matrices (display field of view, 15 to 20 cm; scan field of view, 24 cm). Second, by use of the aforementioned identified slice, dynamic single-slice imaging was performed repeatedly every 4 seconds beginning at the time of contrast medium administration and continuing for 120 seconds. Contrast mediumb (0.21 mL of iohexol/kg [350 mg/mL; 75 mg/kg]) was administered at a constant rate (2 mL/s, IV) by use of a pressure injector.c Third, the cranial portion of the abdomen was scanned again to determine renal volume after contrast enhancement (this examination could be extended to include the caudal portion of the abdomen if a lesion was identified before contrast medium administration). The image acquisition settings were the same for all 3 scans, except for slice thickness, which was increased to 8 mm during dynamic acquisition.

Image analysis was performed by a second-year resident in a veterinary diagnostic imaging program. Analysis included all CT calculations of GFR, which were based on Patlak plot analysis.13–15 An ROI was drawn around each kidney (excluding the renal hilus and main vessels) on each dynamic slice and on the single equivalent slice by use of the precontrast baseline acquisition. The difference between the number of HUs on each dynamic slice and the precontrast baseline slice was used to calculate a corrected kidney HU value. Similar ROIs were centered on the aorta. Corrected aortic HU values were used to approximate the amount of iodine in the blood so that it could be subtracted from the corrected kidney HU values to yield the amount of iodine filtered by the glomeruli. Patlak plot analysis was performed by use of commercial software.d

Renal volume was calculated by use of CT workstation softwaree that segmented each kidney from adjacent structures; segmentation was manually adjusted to ensure that abdominal organs, perihilar fat, and renal blood vessels were excluded from the volume. The slope from Patlak analysis of each kidney was corrected for PCV, multiplied by the renal volume, and then divided by the body weight of the cat to yield GFR of each kidney. The CT-GFR was determined by summing the GFR values for the left and right kidney.13,14 The CT-L-GFR and CT-R-GFR were reported as a percentage of CT-GFR.

Scintigraphy—Functional renal scintigraphy was performed by 2 methods: image-based scintigraphy (for global GFR and GFR of each kidney) and PC-GFR (for global GFR). Image-based renal scintigraphy is a dynamic image-acquisition technique that may be used to estimate global GFR determined via nuclear medicine (scintigraphy) or GFR of each kidney (NM-L-GFR and NM-R-GFR). The GFR determined by use of 99mTc-DTPA plasma clearance measures the rate of radiopharmaceutical clearance from blood, and it is a more valid measure of global GFR than is NM-GFR.7 All scintigraphic calculations of GFR, including drawing of ROIs, were performed by a nuclear-medicine technician who had 8 years of experience. Examiners who performed CT and scintigraphy calculations were unaware of each other's results.

Image-based scintigraphy was performed in accordance with a published protocol16 by use of a gamma camera fitted with a low-energy, all-purpose collimator.f A syringe containing the radiopharmaceutical (approx 2 mCi of 99mTc-DTPA) was prepared. The amount of radioactivity was determined immediately before injection by counting the radioactivity of the unshielded syringe for 60 seconds at a distance of 27 cm from the center of the gamma camera. Prior to injection, each cat was placed in dorsal recumbency with the gamma camera positioned under the table and centered on the cat's kidneys. Dynamic-frame acquisition of the data by use of a 64 × 64 × 16 matrix began immediately before bolus injection of the radiopharmaceutical, which was followed by flushing with saline (0.9% NaCl) solution (2 mL, IV). Data were collected at 6 s/frame for 6 minutes (total of 60 frames). After data collection, the count for the unshielded syringe was conducted again to determine residual radioactivity and establish the net dose administered. By use of nuclear medicine software,g the dynamic scan was summed to a static image of the abdomen (centered on the kidneys), and ROIs were manually drawn around each kidney and background (a crescent-shaped ROI approximately 2 pixels wide at the caudal border of each kidney). Commercially available softwareh was used to calculate the percentage of dose uptake of each kidney minus background. From these data, NM-L-GFR and NM-R-GFR were estimated by use of an established equation built into the software.

The GFR determined by 99mTc-DTPA plasma clearance was obtained from serial blood samples (1.5 mL) collected into heparinized tubes at 15, 30, 60, 150, and 240 minutes after radioisotope injection during scintigraphy.17 Samples were centrifuged and plasma harvested (0.5 mL) for counting in a sodium iodide well counter.i To account for radioactive decay, counting was performed after all 5 samples were collected. A standard also was created to convert the number of millicuries to the number of counts per milliliter. For this standard, the radiopharmaceutical (2 mCi of 99mTc-DTPA) was placed into a volumetric flask, which was then filled to a volume of 1 L with distilled water. By use of commercially available software,j a plasma clearance curve was created by plotting the natural logarithm of the number of counts per milliliter versus time after injection. Linear regression analysis and the area under the curve were used to calculate PC-GFR; these calculations were performed by a board-certified veterinary radiologist.

Statistical analysis—Data were summarized for each method. Comparisons were made between results of PC-GFR and CT-GFR, NM-L-GFR and CT-L-GFR, and NM-R-GFR and CT-R-GFR. These comparisons were made subjectively and objectively. Subjective assessment was performed by creating scatterplots to visually assess the location of the results relative to the line of agreement. Objective assessment was performed by determining the limits of agreement by use of Bland-Altman plots.18,19 Methods were considered interchangeable and acceptable for clinical use when the 95% limits of agreement (mean difference between methods ± 1.96 SDs of the differences) were ≤ 0.7 mL/min/kg. Statistical analyses were performed by use of commercially available computer software.j,k

Results

The global GFR and percentage GFR of the left and right kidneys were tabulated for each method and all 8 cats (Table 1). Comparisons between methods were plotted (Figures 1 and 2). When comparing PC-GFR and CT-GFR, 5 of 8 cats had an absolute difference in global GFR that was < 0.7 mL/min/kg (the limit of agreement defined at the outset of the present study for acceptability in clinical use). The maximum difference (absolute value) was 1.6 mL/min/kg, and the limits of agreement were 1.4 and −1.7 mL/min/kg. The mean ± SD difference was −0.2 ± 0.8 mL/min/kg. When comparing the percentage GFR of the left kidney (NM-L-GFR vs CT-L-GFR), the maximum difference (absolute value) was 20%, the mean ± SD difference was −2.4 ± 10.5%, and the limits of agreement were −22.9% and 18.1%. Conversely, when comparing the percentage GFR of the right kidney (NM-R-GFR vs CT-R-GFR), the maximum difference (absolute value) was 20%, the mean ± SD difference was 2.4 ± 10.5%, and the limits of agreement were −18.1% and 22.9%.

Figure 1—
Figure 1—

Scatterplot (A) and Bland-Altman plot (B) of global GFR assessment determined by use of PC-GFR and CT-GFR in 8 cats. Each square represents results for 1 cat. In the scatterplot, the diagonal line is the line of perfect agreement. In the Bland-Altman plot, the solid horizontal line indicates the mean difference between the methods and the dashed horizontal lines indicate the 95% limits of agreement (ie, ± 1.96 SD).

Citation: American Journal of Veterinary Research 73, 4; 10.2460/ajvr.73.4.463

Figure 2—
Figure 2—

Scatterplots (A and B) and Bland-Altman plots (C and D) of the percentage GFR for the left (A and C) and right (B and D) kidney determined by use of NM-L-GFR and CT-L-GFR in 8 cats. The GFR for each kidney is reported as the percentage of global GFR. See Figure 1 for remainder of key.

Citation: American Journal of Veterinary Research 73, 4; 10.2460/ajvr.73.4.463

Table 1—

Mean ± SD and range values for global GFR and proportional GFR for each kidney determined by use of 3 methods in 8 cats.*

VariableNM-L-GFR (%)NM-R-GFR (%)PC-GFR (mL/min/kg)CT-GFR (mL/min/kg)CT-L-GFR (%)CT-R-GFR (%)
Mean ± SD52.0 ± 0.148.0 ± 0.12.4 ± 0.82.6 ± 1.049.0 ± 0.151.0 ± 0.1
Range46–5941–541.2–3.41.1–4.137–6733–63

Technical difficulties were detected during CT in 2 cats.

For individual kidneys, the GFR is expressed as a percentage of global filtration; therefore, GFR for individual kidneys is the product of the percentage times global GFR.

Technical complications were observed in 2 of 3 cats that had a difference of > 0.7 mL/min/kg in global GFR between methods. One cat moved slightly 84 seconds after start of the dynamic acquisition, which resulted in cranial shifting of the right kidney, so there was less renal volume in the dynamic slices. The corrected HUs of the right kidney ROIs subsequently decreased by > 30% (from 31.4 HUs at 84 seconds to 21.4 HUs at 88 seconds). If GFR was calculated only from 0 to 84 seconds, then CT-GFR was 3.4 mL/min/kg, CT-L-GFR was 48%, and CT-R-GFR was 52%, which were results almost identical to PC-GFR, NM-L-GFR, and NM-R-GFR for this cat. In another cat, the position of the right kidney moved cranially immediately before the start of the dynamic acquisition, and only 7 of 31 ROIs could be drawn. In a clinical setting, the acquisition would be nondiagnostic for CT-R-GFR; however, CT-L-GFR was unaffected in that cat.

Discussion

Whereas the limits of agreement for CT-GFR (compared with those for PC-GFR) exceeded the definition of acceptable clinical use (0.7 mL/min/kg) for the study reported here, the difference between methods was less than this criterion in 5 of 8 cats. Additionally, values that exceeded this criterion in 2 of 3 cats likely did so because of technical complications associated with respiratory motion or slice thickness. Respiratory motion may be diminished in future studies by hyperventilating cats prior to image acquisition. Hyperventilation decreases the carbon dioxide concentration in the blood, which causes a brief period of apnea until the concentration increases enough to stimulate respiration. However, hyperventilation may affect pH and arteriole resistance of the glomeruli, which could alter GFR and act as a confounding factor if the effect of hyperventilation is not consistent between methods. A large slice thickness (8 mm) was used in the present study to maximize the volume of renal parenchyma that was evaluated. A thinner slice (eg, 4 to 5 mm) would allow a kidney to move a few millimeters but potentially retain enough tissue in the dynamic slice to not substantially affect the calculation of GFR. Investigators in a report20 of use of a thinner slice thickness (5 mm) in 4 cats did not compare CT-GFR with other methods of GFR estimation. On the basis of these observations, we concluded that GFR estimation by use of dynamic single-slice CT has potential to be an acceptable clinical technique, and further investigation is needed in a larger sample population with a wide range of GFR values to allow the limits of agreement to be appropriately estimated with a small confidence interval.21

The observed limits of agreement may have been excessively large because of the small sample size, rather than because of differences between methods or technical complications. However, the sample size was sufficient to enable us to provide proof of the concept before enrolling additional cats, including cats with subclinical or overt renal disease, to include a wide spectrum of GFR values. Additionally, the criterion for acceptable clinical use is arbitrary. However, the value (0.7 mL/min/kg) was selected before data were collected and was established on the basis of the clinical experience of the authors. Others may choose a more stringent or less rigorous value. The difference between measurements that do not cause difficulties is a question of judgment.22

The GFR for individual kidneys differed approximately 20% between methods. Criteria for acceptable agreement between methods were not established prior to data collection for percentage excretion by the kidneys because the primary interest was to compare global GFR. However, this variation seems high and may be partially explained by the technical difficulties encountered during measurement for 2 cats, considering that these cats had the greatest differences. Part of this variation may also be explained by differences between methods, which could have a substantial impact on patient care. Because we only compared agreement in results between methods in the present study, another study would be needed to determine the method that is most valid. It is possible that the CT method is more valid because there is lower spatial resolution in the image-based scintigraphic method.

Premedication and anesthetic agents may influence regulation of GFR. We were careful to choose agents that had the least known potential for altering GFR.23,24 We hoped to cancel out potential effects of the agents by performing both examinations (scintigraphy and dynamic CT) immediately after each other and randomizing the order of examinations. However, cats were not anesthetized for the duration of the data collection for plasma clearance (4 hours), and GFR may have been different for blood samples collected throughout the study. In clinical practice, premedication agents might be avoidable when a catheter can be placed in a peripheral vein without the need to sedate the patient. In the present study, an additional catheter was needed for collection of blood samples, so premedication was incorporated into the standard protocol to ensure that all cats received the same drugs and required catheters.

Patient exposure to ionizing radiation is a concern when performing CT examinations. Exposure was minimized by selecting the lowest values for kilovolt (peak) and milliampere allowed by the CT scanner (120 kVp and 50 mA, respectively). Nevertheless, a single slice through the kidneys was repeated > 30 times. Developing protocols that reduce the number of dynamic slices may be warranted as a method for preventing potential problems associated with increased patient exposure.

For administration of contrast medium, we used a low-dose protocol because iohexol is potentially nephrotoxic when used at high doses in human patients with renal disease.25 However, our automated injector only administered volumes in 1-mL increments, so some cats received 1 mL of iohexol, IV, and were underdosed. Additionally, the catheter could not be flushed with saline solution, so part of the small volume of contrast medium likely remained in the catheter in the cephalic vein and did not reach the systemic circulation. Possible remedies include use of a dual auto-injector that would inject iohexol, followed by a saline solution flush. Rather than purchasing a new piece of equipment, dilution of iohexol with saline solution to create a larger bolus for delivery would improve dosing accuracy and increase the amount of contrast medium that reaches the systemic circulation; however, dilution would need to be balanced with the desire to use small volumes to ensure rapid delivery of the contrast medium bolus to the kidneys.

Other methods of CT-GFR calculation exist in addition to the traditional Patlak analysis used in the present study. A common alternate method is modified Patlak analysis. This method differs from traditional Patlak analysis because the ROI is drawn only around cortical tissue, excluding the medulla. Data sampling for the cortical ROI begins with the start of the dynamic scan but must stop when iodinated contrast medium peaks in the proximal tubules (usually < 40 seconds). Thus, more slices are needed early during the dynamic CT scan (interval between slices is 0.5 to 2.5 seconds), which increases the radiation dose unless the interval between slices is greatly increased (6 to 8 seconds) during the later portions of the scan. In 1 study26 in pigs, the modified Patlak method had tighter limits of agreement than the original Patlak method when both methods were compared with inulin clearance.

Contrast-enhanced dynamic MRI is another method for estimation of GFR in humans. Contrast-enhanced dynamic MRI allows estimation of GFR of each kidney from the entire kidney rather than from a partial volume that is collected with CT-GFR methods. Contrast-enhanced dynamic MRI also provides good morphological information without use of ionizing radiation and potentially provides same-day results. The main disadvantages in veterinary medicine are the need for anesthesia, cost of the examination, and availability of a magnetic resonance scanner. Use of gadolinium contrast agents in humans with renal disease increases the risk for development of nephrogenic systemic fibrosis27; however, to our knowledge, this condition has not been reported in domestic animals. Investigators conducted a meta-analysis28 of contrast-enhanced dynamic MRI in humans and concluded that no contrast-enhanced dynamic MRI approach used provided a satisfactory measure of renal function and that GFR results for contrast-enhanced dynamic MRI were not suitable for use as a routine clinical or research technique.

To make dynamic single-slice CT a practical test for real-time clinical estimation of GFR, automated or semiautomated computer algorithms are needed to repeatedly mark ROIs and perform calculations. Calculation of GFR following image acquisition required approximately 90 min/cat.

For the study reported here, we concluded that single-slice dynamic CT has potential for use in the estimation of GFR in cats in a clinical setting. Additional studies are warranted, especially for cats with mild or questionable renal disease, to increase the sample population size and thereby establish a better understanding of the limits of agreement. Avoidable technical complications associated with respiratory motion and slice thickness were also identified in the present study. Development of automated or semiautomated computer algorithms to calculate GFR would improve the clinical applicability of dynamic CT examination.

ABBREVIATIONS

CT

Computed tomography

CT-GFR

Global glomerular filtration rate determined via computed tomography

CT-L-GFR

Percentage of left kidney glomerular filtration rate determined via computed tomography

CT-R-GFR

Percentage of right kidney glomerular filtration rate determined via computed tomography

GFR

Glomerular filtration rate

HU

Hounsfield unit

MRI

Magnetic resonance imaging

NM-L-GFR

Percentage of left kidney glomerular filtration rate determined via nuclear medicine

NM-R-GFR

Percentage of right kidney glomerular filtration rate determined via nuclear medicine

PC-GFR

Global glomerular filtration rate determined via plasma clearance of technetium Tc 99m pentetate

ROI

Region of interest

99mTc-DTPA

Technetium Tc 99m pentetate

a.

Aquilion LB, Toshiba America Medical Systems, Tustin, Calif.

b.

Omnipaque, 350 mg of iodine/mL, GE Healthcare Inc, Princeton, NJ.

c.

MEDRAD Inc, Warrendale, Pa.

d.

PMOD, PMOD Technologies Ltd, Zurich, Switzerland.

e.

Toshiba America Medical Systems, Tustin, Calif.

f.

Ohio-Nuclear Inc, Solon, Ohio.

g.

MEDX Inc, Arlington Heights, Ill.

h.

Alfanuclear, Buenos Aires, Argentina.

i.

Ortec, Oak Ridge, Tenn.

j.

Excel, Microsoft Corp, Redmond, Wash.

k.

MedCalc, version 8.2.1.0, MedCalc Software, Mariakerke, Belgium.

References

  • 1.

    Polzin DJ, Osborne CA & Ross S. Chronic kidney disease. In: Ettinger SJ, Feldman EC, eds. Textbook of veterinary internal medicine. Vol II. 6th ed. St Louis: Elsevier Saunders, 2005;17561785.

    • Search Google Scholar
    • Export Citation
  • 2.

    DiBartola SP. Renal disease: clinical approach and laboratory evaluation. In: Ettinger SJ, Feldman EC, eds. Textbook of veterinary internal medicine. Vol II. 6th ed. St Louis: Elsevier Saunders, 2005;17161730.

    • Search Google Scholar
    • Export Citation
  • 3.

    Muller-Surr R. Radiopharmaceuticals: their intrarenal handling and localization. In: Murray PC, Ell PJ, eds. Nuclear medicine in clinical diagnosis and treatment. Vol I. New York: Churchill Livingston, 2004;195212.

    • Search Google Scholar
    • Export Citation
  • 4.

    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; 65:15021507.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5.

    Bailey DB, Rassnick KM, Dykes NL, et al. Phase I evaluation of carboplatin by use of a dosing strategy based on a targeted area under the platinum concentration-versus-time curve and individual glomerular filtration rate in cats with tumors. Am J Vet Res 2009; 70:770776.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6.

    Bailey DB, Rassnick KM, Prey JD, et al. Evaluation of serum iohexol clearance for use in predicting carboplatin clearance in cats. Am J Vet Res 2009; 70:11351140.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 7.

    Barthez PY, Hornof WJ, Cowgill LD, et al. Comparison between the scintigraphic uptake and plasma clearance of 99mTc-diethylenetriaminepentacetic acid (DTPA) for the evaluation of the glomerular filtration rate in dogs. Vet Radiol Ultrasound 1998; 39:470474.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8.

    Kyles AE, Hardie EM, Wooden BG, et al. Clinical, clinicopathologic, radiographic, and ultrasonographic abnormalities in cats with ureteral calculi: 163 cases (1984–2002). J Am Vet Med Assoc 2005; 226:932936.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9.

    Hackstein N, Cengiz H & Rau W. Contrast media clearance in a single kidney measured on multiphasic helical CT: results in 50 patients without acute renal disorder. AJR Am J Roentgenol 2002; 178:111118.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10.

    Hackstein N, Bauer J, Hauck EW, et al. Measuring single-kidney glomerular filtration rate on single-detector helical CT using a two-point Patlak plot technique in patients with increased interstitial space. AJR Am J Roentgenol 2003; 181:147156.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11.

    Hackstein N, Cengiz H, Rau W, et al. Glomerular filtration rate measured by using triphasic helical CT with a two-point Patlak plot technique. Radiology 2004; 230:221226.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12.

    Tsushima Y, Blomley MJ, Okabe K, et al. Determination of glomerular filtration rate per unit renal volume using computerized tomography: correlation with conventional measures of total and divided renal function. J Urol 2001; 165:382385.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13.

    O'Dell-Anderson KJ, Twardock R, Grimm JB, et al. Determination of glomerular filtration rate in dogs using contrast-enhanced computed tomography. Vet Radiol Ultrasound 2006; 47:127135.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14.

    Alexander K, Del Castillo JR, Ybarra N, et al. Single-slice dynamic computed tomographic determination of glomerular filtration rate by use of Patlak plot analysis in anesthetized pigs. Am J Vet Res 2007; 68:297304.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15.

    Patlak CS, Blasberg RG, Fenstermacher JD. Graphical evaluation of blood-to-brain transfer constants from multiple-time uptake data. J Cereb Blood Flow Metab 1983; 3:17.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16.

    Twardock AR & Bahr A. Renal scintigraphy. In: Daniel GB, Berry CR, eds. Textbook of veterinary nuclear medicine. 2nd ed. Knoxville, Tenn: American College of Veterinary Radiology, 2006;332333.

    • Search Google Scholar
    • Export Citation
  • 17.

    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; 61:280285.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18.

    Altman DG, Bland JM. Measurement in medicine: the analysis of method comparison studies. Statistician 1983; 32:307317.

  • 19.

    Dewitte K, Fierens C, Stöckl D, et al. Application of the Bland-Altman plot for interpretation of method-comparison studies: a critical investigation of its practice. Clin Chem 2002; 48:799801.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20.

    Alexander K, Dunn M, Carmel EN, et al. Clinical application of Patlak plot CT-GFR in animals with upper urinary tract disease. Vet Radiol Ultrasound 2010; 51:421427.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21.

    Altman DG. Some common problems in medical research. In: Practical statistics for medical research. London: Chapman & Hall/CRC, 1997;396439.

    • Search Google Scholar
    • Export Citation
  • 22.

    Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 327:307310.

  • 23.

    Fusellier M, Desfontis J, Madec S, et al. Influence of three anesthetic protocols on glomerular filtration rate in dogs. Am J Vet Res 2007; 68:807811.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24.

    Newell SM, Ko JC, Ginn PE, et al. Effects of three sedative protocols on glomerular filtration rate in clinically normal dogs. Am J Vet Res 1997; 58:446450.

    • Search Google Scholar
    • Export Citation
  • 25.

    Morcos SK. Contrast medium-induced nephrotoxicity. In: Dawson P, Cosgrove DO, Grainger RG, eds. Textbook of contrast media. Oxford, England: Isis Medical Media, 1999;135148.

    • Search Google Scholar
    • Export Citation
  • 26.

    Daghini E, Juillard L, Haas JA, et al. Comparison of mathematic models for assessment of glomerular filtration rate with electron-beam CT in pigs. Radiology 2007; 242:417424.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 27.

    Chen AY, Zirwas MJ, Heffernan MP. Nephrogenic systemic fibrosis: a review. J Drugs Dermatol 2010; 9:829834.

  • 28.

    Mendichovszky I, Pedersen M, Frøkiær J, et al. How accurate is dynamic contrast-enhanced MRI in the assessment of renal glomerular filtration rate? A critical appraisal. J Magn Reson Imaging 2008; 27:925931.

    • Crossref
    • Search Google Scholar
    • Export Citation

Contributor Notes

Supported by the Feline Health Center at Cornell University.

The authors thank Jeffrey Hobbs and Patricia Reynolds for assistance with image and data acquisition.

Address correspondence to Dr. Scrivani (pvs2@cornell.edu).