| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
INNOVATIVE METHODOLOGY
1BioPhysics Assay Laboratory (BioPAL), Incorporated, Worcester; and 2Western New England Renal and Transplant Associates, Springfield, Massachusetts
Submitted 9 June 2008 ; accepted in final form 2 September 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
renal function; chronic kidney disease; transplant
An accurate measurement of GFR is critical for the management of patients with chronic kidney disease (CKD), renal transplant recipients, and the evaluation of potential transplant donors. This diagnostic test is also needed in renal physiology research and in clinical trials of pharmaceutical agents, where accurate assessment of renal function is important. Estimated GFR derived by evaluating the fluctuation of naturally occurring biomarkers have been shown to provide limited diagnostic value for these applications (8, 11–17, 20), as highlighted in a recent systematic review of 23 studies concerning the performance of creatinine-based GFR estimates in kidney transplant recipients (21).
After careful review by the National Kidney Foundation, the direct measurement of GFR remains the "gold standard" for assessing kidney function (14). The diagnostic procedure to accurately measure GFR is simple and straightforward. The procedure requires the administration of a GFR probe followed by timed blood or blood and urine collections. The historical barrier to universal use of a measured GFR test has been its association with radioactivity or the unsuitability of the analytic component of a nonradioactive test precluding its adoption into a routine clinical laboratory due to the requirement for specialized equipment, complex analytical procedures, and slow sample throughput.
In this report, we describe and present the performance characteristics of the FIT-GFR test. Because this diagnostic test has applications in basic research, veterinary medicine and human clinical research, we also evaluate potential interference from different species serum. As part of our initial validation, we present the results of a clinical study that demonstrate the accuracy of the FIT method to measure GFR in humans. The study comprised healthy volunteers and patients with diagnosed CKD many of whom are kidney transplant recipients. The results of this study show that the FIT-GFR method provides an accurate measurement over a wide range of GFR values.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
ELISA FIT-GFR kit. The FIT-GFR kit (BioPAL, Worcester, MA) contains gadopentetate dimeglumine (Gd-DTPA) concentrate, rabbit anti-Gd-DTPA antiserum, HRP-Gd-DTPA conjugate, and a goat anti-rabbit IgG-coated 96-well plate. The composition of each component is described in the package insert of the kit.
FIT-GFR Kit Validation
Sample interference. Sample interference was evaluated for rat, mouse, feline, canine, and human serum. Samples were diluted 1:0, 1:10, 1:20, 1:50, 1:100 into sample diluent. Fifty microliters of diluted sera sample were pipetted into sample wells in triplicate and processed following the standard assay protocol. The resulting optical density values were compared with the mean zero standard value.
Cross-reactivity. The cross-reactivity of a number of chelates with anti-Gd-DTPA antiserum were assessed. Among the compounds tested were gadolinium, DTPA, Gd-DTPA, DOTA, Gd-DOTA, EDTA, Gd-EDTA, citrate, and succinate. Various common drugs, such as aspirin, Tylenol, and ibuprofen were also tested.
Sensitivity. As previously described by Anderson (2), the limit of detection (LOD) and the limit of quantitation (LOQ) were determined by measuring 15 blank replicates of human serum containing no Gd-DTPA at two dilutions (undiluted and 1:300), 15 blank replicates of human urine containing no Gd-DTPA at two dilutions (undiluted and 1:3,000), and 15 blank diluent samples. The standard curve was run in triplicate. The mean value and standard deviation (SD) for each sample set were determined. The LOD was calculated as the means ± 3SD, and the LOQ calculated as the means value ± 10SD.
Accuracy and precision. Intra-assay replicate analysis (n = 10) and interassay replicate analysis (n = 5) at two levels of Gd-DTPA (low 0.007 µg/ml; high 0.065 µg/ml) were determined. The interassay replicates were collected over 5 mo. Accuracy was defined as the range of percentage differences between the means ± 2SD of back-calculated concentrations and real standard values. Intra-assay and interassay precision were expressed as the percent coefficient of variation (CV) of the measured Gd-DTPA concentration, i.e., %CV = 100·SD/mean.
Comparison Study
Experimental protocol. The study was approved by the Internal Review Board and was performed in adherence to the Declaration of Helsinki, and informed consent was obtained from all subjects. The investigation was performed using 20 subjects (14 men), ranging from 30–71 yr of age. Two subjects were healthy volunteers with no known history of renal disease or diabetes. The remaining 18 subjects had diagnosed CKD, wherein 8 were diabetic, 12 received a kidney transplant, and 1 received a liver transplant. All subjects received Lugol's solution before the start of the study to block thyroidal uptake of radioactive 125I. The height and weight of each subject were measured, and subjects were encouraged to drink water throughout the study.
[125I]iodothalamate (10 µCi, Glofil, IsoTex Diagnostics) and gadopentetate dimeglumine (10 µl/kg, 5 µmol/kg, Gd-DTPA, Magnevist, Berlex) were intravenously administrated to each subject. After injection, blood and urine samples were collected, as shown in Fig. 1. The dose of Gd-DTPA used in the study is 10x lower than the recommended dose for magnetic resonance imaging (MRI). The exact time of the injection and each blood and urine collection was recorded, and the volume of each urine collection was measured. Blood was collected in serum separation tubes without anticoagulant and centrifuged within 1 h.
|
ELISA procedure for FIT-GFR.
Standards were prepared using BioPAL's diluent (0.003, 0.01 0.03, 0.1, 0.3 µg/ml). Using the same diluent, serum and urine samples were diluted. It was determined that serum required a dilution of 1:300 and urine samples required a dilution of 1:3,000 to bring the samples within the active range of the standard curve. Fifty microliters of the standard or diluted sample were combined in sequence in goat anti-rabbit IgG-coated plate wells with 50 µl of HRP-Gd-DTPA conjugate containing a yellow dye and 50 µl of rabbit anti-Gd-DTPA containing a blue dye. Upon combination, the wells containing all reagents became green and were incubated for 90 min at 25°C. The wells were washed with a Tween 20 PBS buffer (Elx50, Biotek Instruments, Winooski, VT). Substrate (100 µl) was added to all wells and incubated for 30 min. Stop reagent (100 µl) was then added, and the optical density at 450 nm for each well was recorded. Using software supplied by the plate reader (Multiskan Spectrum, Thermo Election), data from the standards were fit to a four-parameter logistic function. By interpolation, the concentration of Gd-DTPA present in each sample was determined. The analytic procedure required 
3 h to compete.
GFR calculation via UV/P method.
The protocol used in this study provides three UV/P measurements, as follows:
 |
 |
 |
GFR calculation via blood clearance method. A one-compartment blood clearance method was used in this study (3). The concentration of [125I]iodothalamate (cpm/ml) and Gd-DTPA (µg/ml) in each blood sample was plotted as a function of time. The data were fit to a one-exponential decay function, i.e., Y = Be–bX. The function was integrated over the limits zero to infinity to obtain the area under the curve (AUC), i.e., AUC = B/b (mg·min–1·ml–1). The GFR value is then obtained by dividing the administrated dose by the AUC. The GFR (ml/min) is then normalized per 1.73 m2 of BSA.
Sample stability. Blood and urine samples from selected patients were stored at 22, 4, and –20°C. After 1 mo, samples were removed from storage and the FIT analysis was repeated and compared with the original value.
Data analysis. GFR values obtained by [125I]iodothalamate were considered the clinical standard and were directly compared with GFR values obtained by FIT using the analysis of Bland and Altman (4). FIT-GFR values were compared with the clinical standard by assessing the bias of the GFR test (the difference between the GFR test and the clinical standard) and the precision between the GFR test and the clinical standard (r2), with the accuracy expressed as the percentage of the GFRs that fell within 20% of the clinical standard. In addition, Student's t-test was also used to evaluate statistical differences between methods.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
The corresponding LOD for diluent and human serum (undiluted and 1:300) and urine (undiluted and 1:3,000) was 0.0010 µg/ml for each matrix. The LOQ for diluent was 0.0024 µg/ml, for human serum (undiluted and 1:300) were 0.0031 and 0.0020 µg/ml, respectively, and for urine (undiluted and 1:3,000) were 0.0036 and 0.0024 µg/ml, respectively. The presence of human serum or urine did not have a significant effect on the baseline measurement. Moreover, for additional species tested, there was no measurable sample interference observed with serum at any dilution level, including undiluted serum.
The optical density readout and the corresponding Gd-DTPA concentration for the replicates from the intra-assay analysis of the low and high controls are provided in Table 1. The intra-assay %CVs for the low and high controls were 4.3 and 2.3, respectively, and the corresponding measured Gd-DTPA concentrations were 0.0076 ± 0.0003 and 0.0660 ± 0.0015 µg/ml, respectively. For the interassay analysis, the measured Gd-DTPA concentrations for low and high controls were 0.0075 ± 0.0005 and 0.0627 ± 0.0043 µg/ml, respectively, corresponding to interassay %CVs of 6.7 and 6.8, respectively.
|
Comparison Study
For the 20 human subjects tested, the range of GFR values was 10.9–102.1 ml·min–1·1.73 m–2. The FIT-GFR test provides comparable results to those obtained by the FDA-approved test, demonstrating a high degree of precision and accuracy (Figs. 2 and 3). The comparison of the classic UV/P method shows that the bias between the two tests was –2.07 ml·min–1·1.73 m–2, precision was 3.00 ml·min–1·1.73 m–2, and the accuracy was such that 95% of the subjects had FIT-GFRs within 20% of the nuclear test (see Fig. 2B). For the blood clearance method, the bias between the two tests was –4.58 ml·min–1·1.73 m–2, precision was 6.16 ml·min–1·1.73 m–2, and the accuracy was such that 90% of the subjects had FIT-GFRs within 20% of the nuclear blood clearance test (see Fig. 3B). There was no statistical difference measured between FIT-GFRs obtained by either UV/P or the blood clearance method compared with the clinical standard.
|
|
After 1 mo, there was no significant change between the first and second FIT-GFR values observed with samples stored at 22 and 4°C. However, samples stored at –20°C proved to be unstable.
Serum and urine samples that were stored at 4°C were diluted 1:10 with a 0.01 M Tris buffer (pH 7.4) containing 10% glycerol and then stored at –20°C. After 1 mo, there was no significant change between the first FIT-GFR value obtained 2 mo earlier compared the FIT-GFR value obtained from frozen samples stabilized with 10% glycerol (data not shown).
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
This report presents for the first time a technologically realistic approach to measure GFR in the laboratory setting. The analytic arm of this test relies on FIT. This immunoassay-based approach eliminates all major analytic barriers, because most clinical and research laboratories already have the equipment and trained personnel necessary to run an ELISA. In addition, immunoassay offers high sample throughput and potential automation. The GFR probe used for the test is a known GFR marker that is nonradioactive and is a clinically available contrast reagent (10). Given the sensitivity of immunoassay, the dose required for the FIT-GFR test can potentially be orders of magnitude lower than the required dose for a standard MRI examination.
As important, this report also introduces the general concept of FIT, which is a new diagnostic platform. Conventional immunoassay tests are analog in nature (present or absent), confirming the diagnosis of disease or the presence of a foreign substance. FIT diverges from this convention by designing tests to evaluate and gauge the functional performance of an organ or organ system by sequential measurements of an exogenously administrated compound, thereby allowing the calculation of a kinetic parameter (see Fig. 4). FIT combines the techniques of immunoassay as a readout system with the concept of pharmacokinetics, the introduction of a xenobiotic reagent, to measure the kinetic behavior of an administrated compound or set of compounds. Limits on the rate of clearance or accumulation of the compound as measured in biological material provide the input parameters to calculate a functional index for the system of interest. GFR is one example of a functional parameter that can be accurately measured by FIT.
|
0.1 ml) subjected this measurement to a greater degree of error. The poorer result obtained by the blood clearance method is largely due to errors associated with measuring the administrated dose of each compound. Nevertheless, even the results of this comparison are good. At the Gd-DTPA dose used in this study, serum required a dilution of 1:300 and urine samples required a dilution of 1:3,000 to bring the samples within the active range of the standard curve. Because no measurable sample interference was observed with human serum and urine suggests that the administrated dose could be lowered by as much as two orders of magnitude, which in turn would result in a proportional reduction in the dilution factors for serum and urine. In addition, the high level of sensitivity offered by FIT suggests that single-nephron GFR measurements are possible using this technology.
In addition to the ultralow dose of the GFR probe, another advantage of the FIT platform is that it requires only a small amount of sample for processing, i.e., 50 µl of serum or urine or less. As a result, this diagnostic tool can be more easily used to measure GFR in small animal models, such as rodents. Because only a minute amount of sample is consumed by the analysis and because Gd-DTPA is stable in serum and urine samples, samples can be archived and then reassayed at a later date. This is another important advantage for clinical research and drug development. More importantly, because the integrity of the sample does not degrade over time, samples can be collected at a clinical setting and then transported to a remote clinical reference laboratory for analysis. Because Gd-DTPA is a clinically available MRI contrast reagent, the FIT-GFR test offers researchers the ability to use the same diagnostic probe throughout the pharmaceutical development pipeline, i.e., preclinical through human clinical trials.
Measured GFR vs. Estimated GFR
The main advantage of a measured GFR is its accuracy, reproducibility, and reliability. A measured GFR test relies on measuring the renal clearance of an administrated filtration marker under controlled conditions. An ideal filtration marker is a reagent that is freely filtered at the glomerulus, is not bound to plasma proteins, is not reabsorbed, secreted, or metabolized by the tubules, is nontoxic, and is physiologically inert. Therefore, the kinetic clearance of a filtration marker is only influenced by a change in kidney function and will always provide the correct GFR value. As a result, a measured GFR remains the "gold standard" for assessing kidney function (14).
An estimated GFR is derived from evaluating the fluctuation of naturally existing biomarkers. Biomarkers are compounds that exist naturally in the body and can be excreted by the kidney. The accumulation or reduction of these biomarkers in the body can provide an estimate of renal function. Unfortunately, all biomarkers are influenced by a wide range of nonrenal factors that can vary across different patient populations and disease states and can also vary in time, compromising the accuracy of the test. For research applications, biomarkers can vary across different mammalian models. A great deal of effort has been devoted toward the development of correction formulas to improve the diagnostic value of biomarker-based estimates. Although these efforts have attained some level of success, given the large number of variables, both known and unknown, estimated GFR tests are not likely to achieve the high level of accuracy, reproducibility, and reliability associated with a measurement-based test.
Gd-DTPA as a GFR Probe
A major achievement of this report is the demonstration that an ELISA test can be developed for a small-molecular-weight reagent for the purpose of measuring GFR. Although alternative reagents could have been chosen, Gd-DTPA was selected because 1) Gd-DTPA has an excellent safely profile compared with iodinated CT contrast reagents; and unlike inulin, 1) Gd-DTPA is clinically available for human use 3) is not expensive, and 4) is available in more than 100 countries worldwide. As a result, an ELISA-based, Gd-DTPA GFR test could be universally used in most research and clinical settings.
Recently, there have been concerns raised about the general use of gadolinium-containing compounds in patients with severe renal insufficiency, i.e., stage V patients having a GFR of 15 ml·min–1·1.73 m–2 or lower, due to gadolinium's potential association as a trigger for nephrogenic systemic fibrosis (NSF) in this patient subgroup (19). NSF is a systemic fibrosing disorder that was first described in the literature in 2000, with the first reported case going back to 1997 (6). The disease has been reported only in patients with severe renal impairment. The etiology of NSF is still unknown but is likely multifactorial, having a strong link to specific gadolinium-containing contrast reagents.
As of April 2008, the vast majority (91%) of confirmed clinical NSF cases are attributed to the use of one specific gadolinium-containing reagent, gadodiamide (Omniscan; Gd-DTPA-BMA), while in comparison, only 4% of the confirmed cases were associated with Gd-DTPA and 3% of the cases involved patients that were not exposed to gadolinium-containing reagents (5). This disparity cannot be attributed to market share, because Magnevist (Gd-DTPA) outsells Omniscan by nearly 2:1 in the United States (1). The current body of knowledge is largely based on retrospective clinical studies, wherein proposed mechanisms for the root cause cannot be tested. Recently, two basic science investigations have emerged that have tested potential trigger mechanisms under control conditions. Edward et al. (7) showed that the compound Omniscan acts as a growth factor and promotes the growth of fibroblasts. This effect was specific to Omniscan, not free gadolinium, and was dose dependent. Their findings complement a second paper that demonstrates that high doses of Omniscan will induce NSF-like pathology in rats, but high doses of Gd-DTPA will not induce NSF (18). Although the cause of NSF still remains unknown, the link between NSF and gadolinium-containing compounds may prove to be dose dependent and compound specific. Although caution should always be given when regents are administered that clear via the kidneys to patients with severe renal insufficiency, the FIT-GFR test uses an ultralow dose of Gd-DTPA as the GFR probe and is expected to be safe for the vast majority of research applications.
Cost Advantage of FIT-GFR Test
An advantage of the FIT-GFR test over competing analytic methods is cost. The FIT-GFR kit is currently provided as a fixed 96-well plate, wherein 12 wells are needed for the 5 standards and 1 blank that are run in duplicate. Therefore, the cost on a per subject basis is tied to the number of samples generated by the chosen GFR protocol. For example, in the human clinical trial presented above, two different GFR protocols were used. For the UV/P protocol, seven samples needed to be analyzed in duplicate per subject. As a result, six subjects can be processed per plate at a direct cost of $50 per GFR measurement, whereas for the blood clearance protocol, only four samples per subject were needed for analysis to obtain a GFR measurement. Therefore, 10 subjects can be processed per plate at a cost of $30 per GFR measurement. Many researchers advocated the three-point blood clearance method to minimize the number of blood collections. For such a protocol, 14 subjects can be processed per plate, reducing the cost to $21 per GFR measurement. Because only a minute amount of Gd-DTPA (Magnevist) is needed per injection, the probe cost per subject is negligible. Moreover, because most analytic and clinical laboratories routinely run ELISA tests, these facilities already have the equipment, infrastructure, and trained personnel needed to run a FIT test. Therefore, the startup costs are low.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
DISCLOSURES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. A. Stevens and A. S. Levey Measured GFR as a Confirmatory Test for Estimated GFR J. Am. Soc. Nephrol., November 1, 2009; 20(11): 2305 - 2313. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Reinhardt, E. Groman, and D. Vaccaro eGFR--Why read tea leaves when you can make the measurement? Nephrol. Dial. Transplant., June 1, 2009; 24(6): 2002 - 2002. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
