Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print January 5, 2007, 10.1148/radiol.2423060209
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Appendix
Right arrow All Versions of this Article:
2423060209v1
242/3/783    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Boss, A.
Right arrow Articles by Schlemmer, H.-P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Boss, A.
Right arrow Articles by Schlemmer, H.-P.
(Radiology 2007;242:783-790.)
© RSNA, 2007

Genitourinary Imaging

Quantitative Assessment of Glomerular Filtration Rate with MR Gadolinium Slope Clearance Measurements: A Phase I Trial1

Andreas Boss, MD, Petros Martirosian, PhD, Michael Gehrmann, MD, Ferruh Artunc, MD, Teut Risler, MD, Niels Oesingmann, PhD, Claus D. Claussen, MD, Fritz Schick, MD, PhD, Klaus Küper, MD{dagger} and Heinz-Peter Schlemmer, MD

1 From the Section of Experimental Radiology (A.B., P.M., F.S.), Department of Diagnostic Radiology (A.B., M.G., C.D.C., K.K., H.P.S.); and Section of Nephrology and Hypertension, Department of Internal Medicine (F.A., T.R.), Eberhard-Karls University of Tübingen, Hoppe Seyler Strasse 3, 72076 Tübingen, Germany; and Department of Magnetic Resonance, Siemens Medical Solutions, Erlangen, Germany (N.O.). Received February 2, 2006; revision requested March 29; revision received April 18; accepted May 17; final version accepted June 16. Address correspondence to A.B. (e-mail: andreas.boss{at}med.uni-tuebingen.de).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
 References
 
Purpose: To prospectively demonstrate the feasibility of quantifying the glomerular filtration rate (GFR) by assessing the renal clearance of gadolinium-based contrast medium from the extracellular fluid volume in healthy volunteers.

Materials and Methods: The study was approved by the ethics committee and the governmental drug administration department (registration number 4030139, EudraCT number 2004–002969-20, study protocol number 318/2004). Informed consent was obtained from 16 healthy volunteers (six female, 10 male; mean age, 24.5 years ± 2.8 [standard deviation]). Thirteen volunteers (four women, nine men; mean age, 24.8 years ± 2.7; range, 23–30 years) successfully contributed to the study. The GFR was assessed by recording the renal clearance of gadobutrol (3.75 mL, approximately 0.05 mmol per kilogram of body weight) at navigator-gated turbo fast low-angle shot magnetic resonance (MR) imaging. Time–signal intensity curves were constructed from manually drawn regions of interest in the liver, spleen, and renal cortex, and the GFR was calculated by using exponential fitting. Simultaneously obtained iopromide clearance measurements were the reference standard. Statistical evaluations included Bland-Altman plotting and analysis of the relative deviation from iopromide clearance.

Results: Evaluation of liver regions of interest revealed the lowest mean of paired differences from the iopromide clearance measurements (–5.9 mL/min per 1.73 m2 ± 14.6), with a mean GFR of 109.0 mL/min per 1.73 m2 ± 17.1 (134.1 mL/min per 1.73 m2 ± 35.4 for spleen, 100.7 mL/min per 1.73 m2 ± 25.1 for renal cortex) compared with a mean GFR of 103.1 mL/min per 1.73 m2 ± 9.4 measured by using iopromide clearance. The maximum deviation of MR-determined gadobutrol clearance values from iopromide clearance values was 29.2%. The mean disposition half-life of gadobutrol measured in the liver was 83.0 minutes ± 14.2 (72.4 minutes ± 20.2 in spleen, 92.6 minutes ± 23.7 in renal cortex).

Conclusion: The described MR imaging method enables absolute quantification of the GFR after routine contrast material–enhanced MR imaging.

Supplemental material: http://radiology.rsnajnls.org/cgi/content/full/2423060209/DC1

© RSNA, 2007


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
 References
 
Magnetic resonance (MR) imaging with use of the gadolinium chelates gadopentetate dimeglumine and gadobutrol has been proposed for the evaluation of renal function (1–9). Both of these gadolinium-based contrast media have been shown to be distributed predominantly in the extracellular fluid volume (ECFV)—with rapid mixing in the ECFV after intravenous administration—and to have renal clearance with dose-dependent first-order kinetic properties (10–12). Both of these gadolinium chelates are eliminated in an unchanged form through the kidneys by means of glomerular filtration; the amounts eliminated extrarenally are negligible (<0.1%). The average rate of renal clearance of gadobutrol has been demonstrated to be approximately 120 mL/min (12), which is comparable to the rates of renal clearance of other aqueous soluble substances used for glomerular filtration rate (GFR) quantification, such as inulin. More than 50% of the administered gadobutrol dose is eliminated via the urine within 2 hours, and almost all of the injected dose is eliminated within 72 hours after administration. Moreover, both of these gadolinium chelates are well tolerated by patients with impaired renal function (13,14).

Lee et al (15) and Rusinek et al (16) reported performing quantitative MR measurements of the GFR by using a three-dimensional fast low-angle shot (FLASH) gradient-echo sequence after the administration of low concentrations of gadopentetate dimeglumine. The quantification method used by these investigators is based on a technique originally proposed for radionuclide scintigraphy by Gates (17), which correlates the tracer uptake in the renal parenchyma during the first few minutes after administration with the single-kidney filtration capacity. With another method, proposed by Hackstein et al (18,19), a two-compartment model of gadolinium distribution is applied in the kidneys for estimation of the GFR. Both methods offer the advantage of a relatively short measurement time. However, both methods are also challenging to perform and involve time-consuming postprocessing.

The purpose of our study was to prospectively demonstrate the feasibility of quantifying the GFR by assessing the renal clearance of a gadolinium-based contrast medium from the ECFV in healthy volunteers.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
 References
 
Industry Support
The MR (gadobutrol, Gadovist) and radiographic (iopromide, Ultravist) contrast media used in this study were provided by Schering (Berlin, Germany). The study was completely initiated and organized by our radiology department. The authors had full control of the study data and information submitted for publication. No authors were employed by Schering.

Volunteers
Sixteen healthy volunteers (six female, 10 male; mean age, 24.5 years ± 2.8 [standard deviation]) were originally included in this study. In all volunteers, medical history, physical examination results, and urea and creatinine levels were assessed within 1 week before the contrast medium injection. Subjects had to have urea and creatinine levels within normal limits to be included in the study. Hypertension, diabetes, and/or kidney disease was an exclusion criterion. Subjects with low thyroid-stimulating hormone levels and known hypersensitivity to gadolinium-based or iodinated contrast material also were excluded. None of the volunteers took any medications while participating in the study. The volunteers were asked to refrain from eating within 4 hours and to refrain from drinking within 2 hours before the MR measurements. They were allowed to drink small amounts of fluid (up to 200 mL) after MR imaging. The volunteers were monitored under the supervision of a medical doctor for at least 4 hours after the contrast medium injection.

This phase I clinical trial was approved by the local ethics committee and the governmental drug administration department (Federal Institute for Drugs and Medical Devices BfArM, Bonn, Germany; registration number 4030139, EudraCT number 2004–002969-20, study protocol number 318/2004). All volunteers gave written informed consent to participate in the study after being informed of the possible hypersensitivity reaction to the contrast media, which was considered an adverse effect. Hypersensitivity reactions were categorized according to the grading proposed by Brown (20): Mild reactions corresponded to effects on the skin and subcutaneous tissue only; moderate reactions, to respiratory, cardiovascular, or gastrointestinal involvement; and severe reactions, to hypoxia, hypotension, or neurologic compromise. Subjects were told to press the available alarm bell to interrupt the measurement in cases of any mild to severe hypersensitivity reaction.

In each volunteer, the ECFV and body surface area (BSA) were determined by using subject body height (H) and weight (W) (21):

Formula 1(1)

Formula 2(2)

MR System
All MR measurements were performed by using a 1.5-T clinical unit (Sonata; Siemens Medical Solutions, Erlangen, Germany). The gradient system operates with a maximum gradient strength of 40 mT/m and a slew rate of 200 T/m/sec in all three spatial directions. Radiofrequency transmission was performed with a body transmitter. Patients were positioned supine. Anterior and posterior flexible phased-array body coils were applied for signal reception.

Phantom Studies
The increase in longitudinal relaxation rate (1/T1) is directly proportional to the concentration of the contrast medium (c) in the tissue according to the following equation (22):

Formula 3(3)
where R is the specific relaxivity of gadolinium, in (mmol/L)–1 · sec–1, and T1 and T1(c) are the tissue longitudinal relaxation constants before and after contrast medium administration, respectively. The signal intensities derived with the applied turbo FLASH MR sequence were evaluated by using different dilutions of gadobutrol. Phantoms consisting of eight bottles filled with the following concentrations of gadobutrol in 0.9% sodium chloride solution were used: 0, 1.0, 0.75, 0.5, 0.25, 0.1, 0.0625, and 0.016 mmol/L.

The T1 relaxation constants of the dilutions were measured by using an inversion-recovery fast spin-echo MR sequence (7000/9.3 [repetition time msec/echo time msec], bandwidth of 190 Hz/pixel, 5-mm section thickness, 200 x 150-mm field of view, 128 x 96 matrix, fast spin-echo factor of 15) with inversion times of 50, 100, 200, 300, 500, 700, 1000, 2000, 3000, and 5000 msec.

The linear region of the turbo FLASH sequence was determined by using computer software (Microsoft Excel; Microsoft, Redmond, Wash) to compute the linear regression between the signal intensity and the longitudinal relaxation rate with successively higher concentrations (at least three concentrations: 0, 0.016, and 0.0625 mmol/L). T1 values were considered to be within the linear region if all concentrations deviated less than 1% from the line of linear regression. The deviation from linearity for high gadolinium concentrations was measured as the discrepancy between the respective signal intensity and the linear regression line obtained from the gadolinium concentrations in the linear region (with T1 higher than 400 msec). All phantom measurements and data evaluations were performed by one author (A.B.).

Measurement Protocol and Sequences
After a T1-weighted gradient-echo localizer image was obtained, T2-weighted half-Fourier single-shot fast spin-echo data sets (1110/118, pixel bandwidth of 5 mm, 5-mm section thickness, fast spin-echo factor of 77) were acquired in all three orientations to provide an anatomic overview.

For measurement of time–signal intensity curves, a navigator-gated turbo FLASH sequence (498/1.25, 8° flip angle, bandwidth of 500 Hz/pixel, 128 x 128 matrix with 28% oversampling, 3.0 x 3.0-mm in-plane resolution) was used and enabled image recording every 5–7 seconds during continuous breathing with an image acquisition time of 0.6 second. Three subsequent nonselective saturation prepulses lasting 500 µsec followed by a 10 mT/m spoiler gradient of 1 msec were used to achieve strong T1 weighting. The time delay from the last saturation pulse to the center of k-space acquisition was set to 300 msec. The image plane with a 20-mm section thickness was slightly angulated and centrally positioned according to the longest extension of the kidneys. After 20 native images were obtained, a bolus injection of 3.75 mL of gadobutrol (concentration of 1.0 mmol/mL, corresponding to approximately 0.05 mmol of gadobutrol per kilogram of body weight) mixed with 20 mL of iopromide (for simultaneous iopromide clearance determination) was administered at 2 mL/sec and followed by a 20-mL normal saline solution flush. Image recording was continued for an additional 70 minutes. The total duration of the MR examination was approximately 80 minutes.

Total GFR Determined by Using MR Imaging
A biphasic decrease in contrast medium concentration in the body can be assumed. The first phase is due to the mixing of the gadobutrol from the vascular space into the ECFV, whereas the second phase is caused by the clearance of gadobutrol from the ECFV by means of glomerular filtration. The relationship between the time constant of the second exponential phase, {alpha}2 (GFR/ECFV), and the GFR can be deduced from theoretic considerations (Appendix E1 [http://radiology.rsnajnls.org/cgi/content/full/2423060209/DC1]) that lead to the following equation:

Formula 4(4)
where S(t) is the measured MR signal intensity, S(0) is the signal intensity that would be measured directly after injection if instant complete mixing were to occur, Q is the {alpha}2, and t is the time. Equation (4) is valid for the time after complete mixing of the contrast media in the ECFV; therefore, S(0) cannot be directly measured. To evaluate {alpha}2, the curve was fitted to the data at the time after complete mixing. To determine the optimal time for exponential curve fitting, the time spans of 35–60, 40–65, and 45–70 minutes after contrast medium injection were evaluated.

The time–signal intensity curve for the whole-body clearance of gadobutrol was measured by using regions of interest (ROIs) in the parenchyma of the liver, spleen, and kidneys. ROIs were drawn by using the "mean curve" tool in the standard Syngo software of the MR unit, which gives the mean signal intensity within the ROI as a function of time. The organs were outlined in the cortical phase after the gadobutrol injection, and the same ROI was evaluated for the other times (by using the mean curve tool). In the kidneys, only the renal cortex was used for ROI evaluation so as to avoid signal contribution from the contrast medium in the tubuli or the renal pelvis, because the gadobutrol in these structures would have already passed through the filter of the glomeruli at the time of curve fitting; therefore, these renal components needed to be avoided for whole-body clearance measurement. The signal intensities of both kidneys were averaged and weighted according to the sizes of the ROIs.

All ROIs were defined by one investigator (A.B., 3 years experience with renal MR imaging). The {alpha}2 was calculated from the time–signal intensity curves by using a least-squares fit with the nonlinear curve-fitting routine "lsqcurvefit" of the Mathlab programming language (MathWorks, Natick, Mass). The mean areas of the ROIs were 70.2 cm2 ± 16.7 (standard deviation) (780.0 pixels ± 185.6) in the liver, 29.1 cm2 ± 16.8 (323.3 pixels ± 186.7) in the spleen, 12.2 cm2 ± 2.6 (135.6 pixels ± 28.8) in the right kidney, and 13.9 cm2 ± 3.1 (154.4 pixels ± 34.4) in the left kidney.

To calculate the absolute GFR, the {alpha}2 has to be multiplied by the ECFV obtained by using Equation (1). For interindividual comparability, the GFR was normalized to a body surface area of 1.73 m2. The elimination half-life of gadobutrol (t1/2) can be determined from the {alpha}2:

Formula 5(5)

Simultaneous Iopromide Clearance Measurement as Reference Standard
The iopromide clearance measurement method that we used is based on x-ray fluorescence analysis to measure the iodine concentration in the plasma. Use of this method has been shown to yield a good estimate of the GFR (23). In this study, a Renalyzer PRX90 (Provalid, Lund, Sweden), with which two americium 241 sources emit 60-keV photons to excite the iodine in the sample, was used to measure the iodine concentration. The excited iodine subsequently emits characteristic x-rays, which are detected by a sodium iodine semiconductor detector. The amount of emitted characteristic radiation is proportional to the iodine concentration in the plasma sample.

Gadobutrol (3.75 mL) and iopromide (20 mL; iodine concentration, 769 mg/mL) were administered in a contrast medium bolus. Plasma samples were taken by an author (M.G.) 1 hour 30 minutes, 2 hours 15 minutes, 3 hours, and 3 hours 45 minutes after the injection and subsequently analyzed.

Statistical Analyses
Mean values ± standard deviations were calculated from normalized total GFR values, with iopromide clearance measurements as the reference standard, and for each gadobutrol clearance evaluation scheme (ie, evaluation in three organs and exponential fit at three time spans after contrast medium injection). In addition, the mean (± standard deviations) elimination half-life for gadobutrol clearance was determined for each evaluation scheme. To assess the accuracy of the different MR clearance evaluation schemes, we calculated Bland-Altman statistics relating the difference between the GFR measured by using iopromide clearance and that measured by using gadobutrol clearance to the corresponding mean value for each measurement pair. From the Bland-Altman data, we calculated the mean difference and standard deviation between the measurement pairs. In addition, for each evaluation scheme the maximum relative deviation from the iopromide clearance measurement was calculated. One author (A.B.) performed the statistical analyses by using Microsoft Excel 2003 and SPSS for Windows, version 12.0.1 (SPSS, Chicago, Ill), software.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCE IN KNOWLEDGE
 References
 
Phantom Study
For the different concentrations of gadobutrol diluted in 0.9% sodium chloride, the mean specific relaxivity derived by using a least-squares fit and Equation (1) was 4.6 (mmol/L)–1 · sec–1 ± 0.3 at 1.5 T. For this dilution series, a completely linear signal behavior of the turbo FLASH sequence was observed at T1 values down to 400 msec (Fig 1). At a T1 of 260 msec, the deviation from linearity was below 15%. At T1 values below 260 msec, however, conspicuous deviations from linearity were observed.


Figure 1
View larger version (10K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Dependence of signal intensity at turbo FLASH MR imaging (498/1.25/300 [repetition time msec/echo time msec/inversion time msec], 8° flip angle) on tissue T1. Graph shows a completely linear relationship with longitudinal relaxation rates (1/T1) of up to 2.5 sec–1, corresponding to a T1 of 400 msec. With higher longitudinal relaxation rates (smaller T1 values), the signal intensity deviates from linearity. With longitudinal relaxation rates up to 3.8 seconds–1 (T1 of 260 msec), the deviation remains below 15%. a.u. = arbitrary units.

 
To calculate the global GFR, the gadobutrol concentration in the liver parenchyma is recorded after the contrast medium is completely mixed in the ECFV. Because the ECFV corresponds to approximately 30% of the body volume, depending on the subject's height and weight, the highest gadobutrol concentration in the liver parenchyma is approximately 0.17 mmol/L. With a gadobutrol relaxivity constant of 4.6 (mmol/L)–1 · sec–1, an intrinsic liver parenchyma T1 of 586 msec at 1.5 T (24), and a greater than 260 msec intrinsic longitudinal relaxation constant after contrast medium administration, use of Equation (3) yielded a maximum gadobutrol concentration of 0.47 mmol/L with a deviation smaller than 15% from linearity. Therefore, for ROI evaluation in the liver parenchyma with a relatively low mean intrinsic T1 of 581 msec ± 39 (1057 msec ± 42 in spleen, 966 msec ± 58 in renal cortex), a linear relationship between signal behavior and contrast medium concentration could be assumed. The T2* shortening effects of the gadolinium concentration acting on the signal intensity were considered to be negligible owing to the extremely short echo time of 1.25 msec.

Volunteer Examinations and Body Measurements
MR examinations were successfully performed in 13 of the 16 volunteers. One female volunteer had a moderate hypersensitivity reaction—consisting of angioedema, dyspnea, wheezing, nausea, vomiting, chest tightness, and abdominal pain—immediately after the contrast medium injection. Consequently, measurements in this subject had to be discontinued so that she could receive medical treatment according to the standard guidelines for anaphylaxis in radiology (intravenous administration of 400 mg cimetidine, 8 mg dimetindene maleate, and 500 mg prednisolone). This volunteer was admitted to the hospital and completely recovered within 12 hours. The governmental drug administration department was informed the next day. In two other volunteers (one female, one male), quantitative MR evaluation could not be performed owing to body movement. The remaining 13 volunteers (four women, nine men; mean age, 24.8 years ± 2.7; age range, 23–30 years) successfully participated in the study. The mean ECFV for these individuals was 13.8 L ± 2.0, and the mean body surface area calculated by using Equations (1) and (2) was 1.88 m2 ± 0.26.

GFR Results
With use of iopromide clearance—the reference standard—measured normalized GFR values in the 13 healthy volunteers ranged from 88 mL/min per 1.73 m2 to 117 mL/min per 1.73 m2, with a mean GFR of 103.1 mL/min per 1.73 m2 ± 9.4 (Table). The standard deviation of 9.4 mL denotes the physiologic variability in normalized GFRs, which is typically between 90 and 130 mL/min per 1.73 m2.


View this table:
[in this window]
[in a new window]

  		Iopromide Clearance and Gadobutrol Clearance Results

 
The best concordance between gadobutrol clearance and iopromide clearance was observed at ROI evaluation in the liver with the exponential fit performed at 40–65 minutes after the contrast medium injection; this fit resulted in a mean paired difference of –5.9 mL/min per 1.73 m2 ± 14.6 (Table; Figs 2, 3). With this evaluation scheme, a mean GFR of 109.0 mL/min per 1.73 m2 ± 17.1 was measured compared with the mean iopromide clearance rate of 103.1 mL/min per 1.73 m2 ± 9.4. Mean GFRs measured in the spleen and renal cortex were 134.1 mL/min per 1.73 m2 ± 35.4 and 100.7 mL/min per 1.73 m2 ± 25.1, respectively. The mean half-life time of gadobutrol in the ECFV of the liver parenchyma was determined to be 83.0 minutes ± 14.2. Mean half-life times of gadobutrol in the ECFV of the spleen and renal cortex parenchyma were 72.4 minutes ± 20.2 and 92.6 minutes ± 23.7, respectively.


Figure 2A
View larger version (102K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: (a) Slightly oblique coronal T1-weighted turbo FLASH MR image (498/1.25/300, 8° flip angle) shows liver (outlined by solid line), spleen (outlined by dashed line), and kidneys (cortex outlined by dotted lines) in the cortical phase after contrast medium injection. Typical ROIs are drawn for evaluation of time–signal intensity curves. (b) Representative time–signal intensity curves corresponding to ROIs shown in a. Owing to the short intrinsic liver T1 of 586 msec (spleen T1 of 1057 msec, renal cortex T1 of 966 msec), the curve for the liver parenchyma shows the highest signal yield. The variability in signal intensity course is caused by thermal noise and small variations in the diaphragm position. Evaluation of the liver parenchyma revealed the smoothest curve. a.u. = arbitrary units. (c) Exponential decay on the logarithmically scaled time–signal intensity curve (displayed for liver parenchyma) indicates linear behavior. Two exponential phases can be distinguished owing to different slopes: the fast mixing phase immediately after contrast medium injection and the more prolonged phase, in which the signal intensity decrease is caused solely by glomerular filtration. The negative slope of this phase is equal to the GFR divided by the ECFV.

 

Figure 2B
View larger version (22K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: (a) Slightly oblique coronal T1-weighted turbo FLASH MR image (498/1.25/300, 8° flip angle) shows liver (outlined by solid line), spleen (outlined by dashed line), and kidneys (cortex outlined by dotted lines) in the cortical phase after contrast medium injection. Typical ROIs are drawn for evaluation of time–signal intensity curves. (b) Representative time–signal intensity curves corresponding to ROIs shown in a. Owing to the short intrinsic liver T1 of 586 msec (spleen T1 of 1057 msec, renal cortex T1 of 966 msec), the curve for the liver parenchyma shows the highest signal yield. The variability in signal intensity course is caused by thermal noise and small variations in the diaphragm position. Evaluation of the liver parenchyma revealed the smoothest curve. a.u. = arbitrary units. (c) Exponential decay on the logarithmically scaled time–signal intensity curve (displayed for liver parenchyma) indicates linear behavior. Two exponential phases can be distinguished owing to different slopes: the fast mixing phase immediately after contrast medium injection and the more prolonged phase, in which the signal intensity decrease is caused solely by glomerular filtration. The negative slope of this phase is equal to the GFR divided by the ECFV.

 

Figure 2C
View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c: (a) Slightly oblique coronal T1-weighted turbo FLASH MR image (498/1.25/300, 8° flip angle) shows liver (outlined by solid line), spleen (outlined by dashed line), and kidneys (cortex outlined by dotted lines) in the cortical phase after contrast medium injection. Typical ROIs are drawn for evaluation of time–signal intensity curves. (b) Representative time–signal intensity curves corresponding to ROIs shown in a. Owing to the short intrinsic liver T1 of 586 msec (spleen T1 of 1057 msec, renal cortex T1 of 966 msec), the curve for the liver parenchyma shows the highest signal yield. The variability in signal intensity course is caused by thermal noise and small variations in the diaphragm position. Evaluation of the liver parenchyma revealed the smoothest curve. a.u. = arbitrary units. (c) Exponential decay on the logarithmically scaled time–signal intensity curve (displayed for liver parenchyma) indicates linear behavior. Two exponential phases can be distinguished owing to different slopes: the fast mixing phase immediately after contrast medium injection and the more prolonged phase, in which the signal intensity decrease is caused solely by glomerular filtration. The negative slope of this phase is equal to the GFR divided by the ECFV.

 

Figure 3A
View larger version (9K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a: Normalized GFR measured by using gadobutrol clearance for MR evaluation of an ROI in the liver at 40–65 minutes after contrast medium injection is compared with iopromide clearance. (a) Graph shows a plot of the GFR values obtained by using both methods. (b) Bland-Altman plot relates the difference between each measurement pair to the mean value. A mean difference in measurement pairs of –5.9 mL/min ± 14.6 was calculated. All measurement points are within ±2 standard deviations from the mean value.

 

Figure 3B
View larger version (10K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b: Normalized GFR measured by using gadobutrol clearance for MR evaluation of an ROI in the liver at 40–65 minutes after contrast medium injection is compared with iopromide clearance. (a) Graph shows a plot of the GFR values obtained by using both methods. (b) Bland-Altman plot relates the difference between each measurement pair to the mean value. A mean difference in measurement pairs of –5.9 mL/min ± 14.6 was calculated. All measurement points are within ±2 standard deviations from the mean value.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCE IN KNOWLEDGE
 References
 
We have described a proposed technique for determining the global GFR by measuring the slope clearance of gadobutrol from the ECFV. The biexponential decrease in signal intensity after bolus gadobutrol injection can be quantified by using two specific time constants: {alpha}1 and {alpha}2. The time constant of the first exponential phase, {alpha}1, is due to the mixing of intravascular contrast medium in the ECFV. The {alpha}2 time constant is equal to the GFR divided by the ECFV (Appendix E1 [http://radiology.rsnajnls.org/cgi/content/full/2423060209/DC1]). We observed good correspondence between MR clearance measurement and iopromide clearance measurement, the reference standard, with a mean paired difference between these measurements of –5.9 mL/min per 1.73 m2 ± 14.6. The half-life disposition times in the volunteers measured with MR imaging corresponded well to reported plasma half-life times measured with atomic emission spectroscopy (12).

To our knowledge, the described method of estimating the global GFR by recording time–signal intensity curves evaluated in organ parenchyma to measure the renal clearance of an MR contrast medium from the ECFV had not been previously described. Three organs were evaluated for measurement of contrast medium clearance from the ECFV: the liver, the spleen, and the renal cortex. The time–signal intensity curves for the liver were found to be the most suitable for determining the GFR, and this may have been owing to the higher intrinsic signal intensity in this organ (due to the short T1 in liver, 586 msec [24], compared with the T1 values in the spleen [1057 msec] and renal cortex [966 msec]) and the lower signal intensity variability due to the larger ROI.

A suitable time span for the least-squares fit to the exponential decrease in time–signal intensity curves had to be chosen. Bland-Altman tests were performed with three time spans. Forty to 65 minutes after contrast medium injection was found to be the best span scheme. Starting times earlier than 40 minutes after injection lead to a substantial contribution of mixing effects to the {alpha}2 and in turn lead to an overestimation of the GFR, because {alpha}1 is much smaller than {alpha}2. Evaluation of the exponential decrease in signal intensity at 35–60 minutes after contrast medium injection led to an overestimation of the GFR by a mean paired difference of –21.4 mL/min per 1.73 m2 ± 23.3. Sixty-five minutes after injection, the time–signal intensity curves had declined to the order of the noise level. Curve fitting that included images acquired later than 65 minutes after injection could not be performed reliably; consequently, the mean paired difference was –3.0 mL/min per 1.73 m2 ± 42.1.

Lee et al (15) described another MR method for computing the absolute GFR, in which the initial uptake of gadolinium in the renal parenchyma is related to kidney function. This method offers the advantage of direct measurement of the absolute single-kidney GFR. In addition, the GFR can be assessed within a relatively short time, as only the first 4–5 minutes after injection have to be recorded. However, the technical requirements are demanding: The conversion function for the signal has to be known for the specific MR sequence and imaging unit. Also, the three-dimensional segmentation is time consuming, requiring approximately 2–3 hours of postprocessing.

Hackstein et al (18,19) described yet another MR method for single-kidney GFR determination, which involved the use of a two-point Patlak plot technique. Excellent correlation with the reference standard, iopromide clearance, was observed; the Pearson correlation coefficient was 0.83. However, this technique has some limitations: The contrast medium distribution in the renal interstitium as a third compartment is not considered, and this may lead to an underestimation of the GFR. Furthermore, this technique, similar to that described by Lee et al, necessitates three-dimensional segmentation.

Being a gadolinium-based contrast medium, gadobutrol was chosen for our study because of its lower osmolarity, as the diuretic effect of gadolinium chelates may influence the GFR. Alternatively, other gadolinium chelates that exhibit predominant distribution in the ECFV, have first-order kinetic properties of renal clearance, and are well tolerated by patients with impaired renal function may be used. However, these agents may have a greater influence on the GFR. Gadopentetate dimeglumine also has been reported to have the necessary contrast medium distribution and clearance characteristics (10,11,13).

The following limitations regarding the described model for quantifying GFR by measuring the renal clearance from the ECFV have to be considered: First, the {alpha}2 is related to the GFR; however, during the 40–65-minute period after contrast medium injection for evaluation of the second phase, a further distribution of gadobutrol in the ECFV takes place with a distribution half-life of approximately 0.2 hour, and this influences the second phase of the exponential decrease in signal intensity. Second, contrast medium excretion occurs during the first phase of the biexponential decline as well and leads to a reduction of available contrast material at the time of exponential fitting. Third, the extrarenal clearance, which accounts for 0.1% of the total clearance by biliary excretion, is not considered in this model. This may lead to inaccuracies, especially in liver ROI evaluations. Fourth, with this approach, the measuring time in the imaging unit is higher than that with the approaches proposed by Lee et al (15), Rusinek et al (16), and Hackstein et al (18,19). Fifth, the error introduced by manually drawing ROIs and automatically copying the ROIs to images at other times could not be quantified in this study. Sixth, the variability in GFR due to the different physiologic states at different times could hamper the use of the reference standard, and the diuretic effect of gadobutrol could lead to overestimation of the GFR. Seventh, minimal body movement during MR evaluation is critical, as the signal intensities need to be measured at the exact same position for reliable curve fitting. Two of the 16 volunteers had to be excluded from the study owing to body movement.

In conclusion, the described MR method enables absolute quantification of the GFR after routine contrast material–enhanced MR imaging. Thus, this method can be combined with contrast-enhanced MR assessment of split renal function. The GFR may also be assessed by measuring the clearance of the already administered contrast medium—especially after conventional contrast-enhanced MR angiography of the renal arteries. Further attempts to increase the signal-to-noise ratio, optimizing the amount of contrast medium used and varying the sequence types, should be undertaken. As the results of this phase I trial are very promising, a phase II trial involving patients with reduced renal clearance capability seems warranted.


   ADVANCE IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
References
 


   FOOTNOTES
 

Abbreviations: ECFV = extracellular fluid volume • FLASH = fast low-angle shot • GFR = glomerular filtration rate • ROI = region of interest

{dagger} Deceased. Back

Author contributions: Guarantors of integrity of entire study, A.B., C.D.C., F.S., K.K., H.P.S.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, A.B., P.M., F.S., K.K., H.P.S.; clinical studies, A.B., M.G., F.A., T.R., F.S., K.K., H.P.S.; experimental studies, A.B., P.M., N.O., F.S.; statistical analysis, A.B., P.M., F.S., K.K., H.P.S.; and manuscript editing, A.B., P.M., C.D.C., F.S., K.K., H.P.S.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
 References
 

  1. Choyke PL, Frank JA, Girton ME, et al. Dynamic Gd-DTPA–enhanced MR imaging of the kidney: experimental results. Radiology 1989;170:713–720.[Abstract/Free Full Text]
  2. Kikinis R, von Schulthess GK, Jager P, et al. Normal and hydronephrotic kidney: evaluation of renal function with contrast-enhanced MR imaging. Radiology 1987;165:837–842.[Abstract/Free Full Text]
  3. Knesplova L, Krestin GP. Magnetic resonance in the assessment of renal function. Eur Radiol 1998;8:201–211.[CrossRef][Medline]
  4. Fukuda Y, Watanabe H, Tomita T, Katayama H, Miyano T, Yabuta K. Evaluation of glomerular function in individual kidneys using dynamic magnetic resonance imaging. Pediatr Radiol 1996;26:324–328.[CrossRef][Medline]
  5. Laissy JP, Faraggi M, Lebtahi R, et al. Functional evaluation of normal and ischemic kidney by means of gadolinium-DOTA enhanced TurboFLASH MR imaging: a preliminary comparison with 99Tc-MAG3 dynamic scintigraphy. Magn Reson Imaging 1994;12:413–419.[CrossRef][Medline]
  6. Pettigrew RI, Avruch L, Dannels W, Coumans J, Bernardino ME. Fast-field-echo MR imaging with Gd-DTPA: physiologic evaluation of the kidney and liver. Radiology 1986;160:561–563.[Abstract/Free Full Text]
  7. Semelka RC, Hricak H, Tomei E, Floth A, Stoller M. Obstructive nephropathy: evaluation with dynamic Gd-DTPA–enhanced MR imaging. Radiology 1990;175:797–803.[Abstract/Free Full Text]
  8. Rohrschneider WK, Haufe S, Wiesel M, et al. Functional and morphologic evaluation of congenital urinary tract dilatation by using combined static-dynamic MR urography: findings in kidneys with a single collecting system. Radiology 2002;224:683–694.[Abstract/Free Full Text]
  9. Teh HS, Ang ES, Wong WC, et al. MR renography using a dynamic gradient-echo sequence and low-dose gadopentetate dimeglumine as an alternative to radionuclide renography. AJR Am J Roentgenol 2003;181:441–450.[Abstract/Free Full Text]
  10. Schuhmann-Giampieri G, Krestin G. Pharmacokinetics of Gd-DTPA in patients with chronic renal failure. Invest Radiol 1991;26:975–979.[Medline]
  11. Krahe T, Gotz R, Lackner K, et al. Pharmacokinetics of gadolinium-DTPA in chronic renal insufficiency requiring dialysis [in German]. Rofo 1992;156:523–526.[Medline]
  12. Staks T, Schuhmann-Giampieri G, Frenzel T, Weinmann HJ, Lange L, Platzek J. Pharmacokinetics, dose proportionality, and tolerability of gadobutrol after single intravenous injection in healthy volunteers. Invest Radiol 1994;29:709–715.[CrossRef][Medline]
  13. Haustein J, Niendorf HP, Krestin G, et al. Renal tolerance of gadolinium-DTPA/dimeglumine in patients with chronic renal failure. Invest Radiol 1992;27:153–156.[CrossRef][Medline]
  14. Tombach B, Bremer C, Reimer P, et al. Renal tolerance of a neutral gadolinium chelate (gadobutrol) in patients with chronic renal failure: results of a randomized study. Radiology 2001;218:651–657.[Abstract/Free Full Text]
  15. Lee VS, Rusinek H, Noz ME, Lee P, Raghavan M, Kramer EL. Dynamic three-dimensional MR renography for the measurement of single kidney function: initial experience. Radiology 2003;227:289–294.[Abstract/Free Full Text]
  16. Rusinek H, Lee VS, Johnson G. Optimal dose of Gd-DTPA in dynamic MR studies. Magn Reson Med 2001;46:312–316.[CrossRef][Medline]
  17. Gates GF. Glomerular filtration rate: estimation from fractional accumulation of Tc-99m DTPA (stannous). AJR Am J Roentgenol 1982;138:565–570.[Abstract/Free Full Text]
  18. Hackstein N, Wiegand C, Rau WS, Langheinrich AC. Glomerular filtration rate measured by using triphasic helical CT with a two-point Patlak plot technique. Radiology 2004;230:221–226.[Abstract/Free Full Text]
  19. Hackstein N, Kooijman H, Tomaselli S, Rau WS. Glomerular filtration rate measured using the Patlak plot technique and contrast-enhanced dynamic MRI with different amounts of gadolinium-DTPA. J Magn Reson Imaging 2005;22:406–414.[CrossRef][Medline]
  20. Brown SG. Clinical features and severity grading of anaphylaxis. J Allergy Clin Immunol 2004;114:371–376.[CrossRef][Medline]
  21. Peters AM. The kinetic basis of glomerular filtration rate measurement and new concepts of indexation to body size. Eur J Nucl Med Mol Imaging 2004;31:137–149.[CrossRef][Medline]
  22. Haacke EM, Brown RW, Thompson MR, Venkatesan R. Magnetic resonance imaging: physical principles and sequence design. New York, NY: Wiley-Liss, 1999; 752.
  23. Gaspari F, Perico N, Ruggenenti P, et al. Plasma clearance of nonradioactive iohexol as a measure of glomerular filtration rate. J Am Soc Nephrol 1995;6:257–263.[Abstract]
  24. de Bazelaire CM, Duhamel GD, Rofsky NM, Alsop DC. MR imaging relaxation times of abdominal and pelvic tissues measured in vivo at 3.0 T: preliminary results. Radiology 2004;230:652–659.[Abstract/Free Full Text]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Appendix
Right arrow All Versions of this Article:
2423060209v1
242/3/783    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Boss, A.
Right arrow Articles by Schlemmer, H.-P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Boss, A.
Right arrow Articles by Schlemmer, H.-P.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE