| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Changi General Hospital, 2, Simei St. 3, S, 529889
Singapore.
2 Department of Nuclear Medicine, Singapore General Hospital, Outram Rd.,
S169610, Singapore.
3 Biostatistics Unit, Division of Clinical Trials and Epidemiological Sciences,
National Cancer Centre Singapore, 11, Hospital Dr., 169610, Singapore.
4 Department of Diagnostic Radiology, Tan Tock Seng Hospital, 11 Jln Tan Tock
Seng, S308433, Singapore.
Received November 22, 2002;
accepted after revision February 7, 2003.
Address correspondence to H. S. Teh.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Patients who underwent both radionuclide and MR renographic studies within a 3-month period were included in the analysis. This yielded 21 studies from 19 patients. Nineteen of the 21 studies were available for analysis. Two studies were excluded because of technical errors during MR renographic acquisition. Serial MR renograms were obtained using a dynamic two-dimensional spoiled gradient-echo fast low-angle shot T1-weighted sequence. Low-dose IV furosemide and gadopentetate dimeglumine (0.025 mmol/kg of body weight) were administered. Intensity–time curves were obtained from the manually selected regions of interest over the renal parenchyma and whole kidney for calculation of split renal function and assessment of urinary excretion, respectively. Results were compared with those obtained with radionuclide renography.
RESULTS. Good correlation (Pearson's correlation coefficient,
r = 0.97, p < 0.001) was observed when the
volume-corrected split renal function acquired with MR renography was compared
with that obtained with radionuclide renography. There was also good agreement
in the excretory curve patterns (weighted 
observer 1 = 0.77
and observer 2 = 0.81) between the two techniques.
CONCLUSION. Dynamic MR gradient-echo imaging with a low-dose gadopentetate dimeglumine technique can produce an intensity–time curve and serial dynamic images of the urinary system, in a way similar to that of radionuclide renography. This technique allows assessment of split renal function and urinary excretory status and is a feasible alternative to radionuclide renography.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
More recently, MR imaging using chelated gadolinium–based agents was described for the assessment of renal function and evaluation of urinary obstruction [4–9]. However, few researchers can produce a renographic curve and dynamic serial images that are comparable to those obtained with radionuclide renography, allowing evaluation of split renal function and urinary excretion [10–12].
The purpose of this study was to evaluate the feasibility of acquiring an MR signal intensity–time renographic curve and dynamic serial images in a way similar to that of acquiring radionuclide renograms, with a dynamic gradient-echo sequence and a low-dose gadopentetate dimeglumine technique, using a commonly available 1.5-T MR scanner (Magnetom Vision, Siemens Medical Systems, Erlangen, Germany).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Patient Preparation
Before the examination, patients were given 3 c (
750 mL) of water
orally to ensure adequate hydration. Patients were also asked to void before
the examination to prevent a full bladder effect, which may distort the
excretory segment of the renographic curve
[4,
10,
13]. Voiding would also
prevent patient discomfort toward the end of the examination due to a full
bladder. No external compression was applied. Low-dose IV furosemide of 0.3
mg/kg of body weight to a maximum of 20 mg ([20 mg/2 mL] Lasix, Novate, Milan,
Italy) was administered 15 min before injection of gadopentetate dimeglumine
(0.025 mmol/kg of body weight). The mean volume of injected gadopentetate
dimeglumine was 3.0 mL (range, 2.0–5.0 mL). Administration of
gadopentetate dimeglumine was followed by 20 mL of physiologic saline to
create a bolus. An injector was used when available (n = 11)
(Spectris MR injector, Medrad, Indianola, PA). The injection rate of
gadopentetate dimeglumine through the injector was 2 mL/min. Patients were
allowed to take shallow breaths. Imaging was performed with a Magnetom Vision
1.5-T MR scanner (Siemens Medical Systems). We imaged patients in a supine
position, using a phased array body surface coil.
Scanning Protocol
The study began with a T1-weighted localizer in three orthogonal planes,
followed by a T2-weighted axial half-Fourier single-shot turbo spin-echo
acquisition (time of acquisition, < 20 sec; TR/TE, 4.4/64.0; slice
thickness, 8.0 mm) and sagittal true fast imaging with steady-state precession
sequences over the kidneys. The latter two sequences allow better
visualization of the renal hilum because of the high signal from the urine in
the renal pelvis. This high signal helps in the selection of the position and
angulation of the coronal images for the subsequent contrast-enhanced dynamic
renographic studies, which were performed during suspended respiration. The
axial images are also used for the computation of total renal volume and to
detect any significant focal renal abnormality that may interfere with the
calculation of the split renal function.
The serial renographic images were obtained using a dynamic two-dimensional spoiled gradient-echo fast low-angle shot T1-weighted sequence (time of acquisition, 5.6 sec; 21.9/4.2; slice thickness, 10.0 mm; flip angle, 60°). The images were acquired in an angulated coronal plane of 10-mm thickness along the long axis of the kidneys. The plane of section was through the renal hilum. The field of view of 340–500 mm was modified to a rectangular configuration and optimized to accommodate the patient's body habitus. A 256 x 256 image matrix was used. All patients were given 0.025 mmol kg–1 IV gadopentetate dimeglumine (Omniscan [0.5 mmol/mL], Nycomed, Oslo, Norway [n = 7]; Dotarem [0.5 mmol/mL], Guerbet, Aulnaysous-Bois, France [n = 2]; Magnevist [0.5 mmol/mL] Schering, Berlin, Germany [n = 12]).
In the initial three studies, images were acquired at 20-sec intervals from the beginning of one acquisition to the start of the next acquisition. This temporal resolution resulted in a failure to acquire the true peak signal intensity in the third study, which was excluded from subsequent analysis. Subsequent studies were performed at a scan interval of 10 sec. Scans were obtained over a period of 20 min. Intensity–time curves were obtained from the manually selected region-of-interest (ROI) over the renal parenchyma and whole kidney (parenchyma and the renal pelvis) for calculation of split renal function and assessment of urinary excretion, respectively (Fig. 1A, 1B, 1C, 1D). The ROI was automatically overlaid for the rest of the series. The ROI position was manually corrected when it was noted to extend significantly beyond the kidney. To correct for possible variation in MR signal intensity between the kidneys of each patient contributed by factors such as difference in patient size and coil positioning, the values obtained (Sx) were corrected relative to the baseline value (So) for each kidney, before contrast administration according to the calculation (Sx – So). These relative signal intensity values were plotted against time and used for all analyses.
|
|
|
|
The renogram typically consists of vascular, parenchymal, and excretory phases (Fig. 2A, 2B). The vascular phase occurs almost immediately after contrast injection and gives the first segment of the renographic curve, which is reflected by a steep linear rise. This is followed by the parenchymal phase, characterized by continuous uptake of contrast material by the whole kidney, which is represented in the intensity–time curve as a slower linear increase up to a second peak. In the excretory phase, contrast material is released into the collecting system calices and constitutes the third segment of the renographic curve.
|
|
Functional Evaluation
Split renal function was determined by calculating the area under the curve
taken from the 1st min of the second segment of the signal
intensity–time curve generated from an ROI drawn over the renal
parenchyma. This calculation gives the mean signal intensity per voxel of
tissue and determines the function per unit volume of the kidney because the
signals acquired are from a single slice of kidney through the renal hilum.
This is in contrast to radionuclide renography in which the radiotracer count
represents the total radiotracer uptake of the whole kidney
[12]. The overall
volume-corrected split renal function ratio between the left and right kidneys
is determined with the following formula: area under the curve x total
volume of left kidney or area under the curve x total volume of right
kidney. Volume calculation was performed using the voxel count also known as
the slice summation method. The outline of each kidney was drawn manually. The
total renal volume was determined from the sum of voxel volumes lying within
the boundaries. This calculation was performed by multiplying the areas within
the segmented sections by the section thickness, including the intersection
gap [14].
Evaluation of Urinary Excretion
The excretory pattern was assessed using the third segments of the
intensity–time renographic curves obtained with the whole-kidney ROI.
The interpretation of the renographic curve was aided by visual analysis of
the dynamic images when necessary. The results obtained by MR renographic
imaging and radionuclide renographic examinations were graded by two
independent observers who were unaware of the other's assessments. Each
diagnostic technique was assessed in isolation by the observers, who were
unaware of the outcome of the other examination. The excretory patterns were
classified as follows: I, prompt and concave fall in the third segment of the
renographic curve; II, obstructive pattern in which an accumulation curve was
observed indicating no washout; IIIa, dilated but nonobstructed pattern in
which there was an initial increase but subsequent complete washout; IIIb,
equivocal borderline obstruction in which a concave excretory pattern was not
observed because of incomplete washout; IV, poor or no renal function with
flattened excretion [1].
Diuretic Renal Scintigraphy
Patients were given oral hydration and asked to empty their bladders before
the procedure. IV furosemide (Lasix, Novate) at a dose of 40 mg was
administered 15 min before the radiotracer injection. For the examination,
99mTc mercaptoacetyltriglycine was used at a dose of 10 µCi. The
patient was imaged in the supine position with his or her back to the camera.
Radiotracer counts were acquired with a large–field-of-view gamma camera
(Sopha Medical Vision, DST-XLi, Buc, France) equipped with a low-energy
all-purpose collimator, which was placed posterior to the patient, using a
photo peak for 99mTc (140 keV) and a 20% window setting. A field of
view of 64 x 64 matrices was used. Images were acquired at 15 sec/frame
for 96 frames. The dynamic study was reframed into 90-sec images for display.
Acquisition occurred over a period of 24 min. Time–activity curves were
generated using a whole-kidney ROI. Background subtraction was obtained with a
crescent-shaped perirenal ROI. For patients with hydronephrosis, an additional
time–activity curve generated from the ROI over the renal parenchyma,
not including the dilated pelvic–caliceal system, was obtained for
determination of the differential function. The differential renal uptake
ratio of 99mTc mercaptoacetyltriglycine was measured at
1.5–2.5 min after injection of the radiopharmaceutical.
Data and Statistical Analysis
Data were transferred to a personal computer for postprocessing using
java-based Digital Imaging and COmmunications in Medicine software (Image J,
Wayne Rasband, Research Services Branch, National Institute of Mental Health,
Bethesda, MD). All statistical analyses were carried out using the Analyse-it
software (version 1.62, Leeds, England, United Kingdom) and SPSS version 10.0
(Statistical Package for the Social Sciences, Chicago, IL). The correlation
and agreement for the volume-corrected single-kidney split renal function
between the MR renogram and the radionuclide renogram were determined. The
correlation measured the strength of the linear association between the two
methods. The Pearson's correlation coefficient was used as the measure of
correlation. The agreement was assessed using the Bland-Altman approach, in
which the bias and 95% limits of agreement between the two methods were
measured [15].
For assessing the agreement in the excretory curve patterns between the two techniques, we made use of the weighted kappa statistic, which assessed the level of agreement between methods with ordinal data. A kappa value greater than 0.80 was deemed as excellent agreement; between 0.61 and 0.80, good agreement; between 0.41 and 0.60, moderate agreement; between 0.21 and 0.40, as fair agreement; equal to or less than 0.20, poor agreement [16].
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
stent in situ, which caused
magnetic susceptibility artifact affecting signal acquisition. In the second
patient, the true parenchymal peak could not be obtained because of suboptimal
temporal resolution when MR renography was performed with a scan time interval
of 20 sec.
Split Renal Function
Of the remaining 19 studies, 18 studies from 16 patients were available for
the evaluation of split renal function. One patient had a single kidney, and
hence split renal function could not be computed. The correlation for the
single kidney split renal function was found to be 0.97 (p <
0.001) (Fig. 3). This indicated
a high level of linear association for the volume-corrected split renal
function between the MR renography and radionuclide renography. As for the
level of agreement, we found a bias of –1.3% in favor of radionuclide
renography (i.e., the radionuclide renographic interpretations were 1.3% lower
than the MR renographic interpretations on average). This was close to zero,
suggesting that the two methods agreed well on average. However, the 95%
limits of agreement ranged from –11.7% to 9.1% This meant that for a new
subject, we expected the two methods to give measurements of a single kidney
split renal function that could differ by as much as approximately 10%. This
was observed in three of our studies. All three patients were noted to have a
hydronephrotic kidney with significantly thinned cortex.
|
Urinary Excretion
Thirty-seven pairs of MR renographic and radionuclide renographic curves
from 19 studies were available for evaluation. Results of MR renographic and
radionuclide excretory renographic curve patterns from the two observers are
summarized in Table 1. With
regard to the agreement in the excretory curve patterns between the two
techniques, weighted kappa statistics measuring agreement of 0.77 and 0.81
were obtained for observer 1 and observer 2, respectively, indicating good
agreement between the two techniques. The interobserver agreement using
weighted kappa statistics for radionuclide renography and MR renography was
0.75 and 0.74, respectively.
|
Several observations were also made from the dynamic serial MR renography (Fig. 4A, 4B, 4C). Before contrast administration, the kidneys showed low-T1 signal intensity. The cortex appeared slightly higher in signal compared to the medulla. The vascular phase occurred almost instantaneously after contrast injection and reached a peak at approximately 23.4 ± 2.0 sec. The renal cortex showed a peripheral rim of high signal intensity. The medulla and the renal pelvis remained dark at this time. This change was caused by the arrival of the contrast bolus in the arcuate, the interlobular arteries, and the glomerular capillaries. This was followed by the parenchymal phase, which showed progressive and centripetal flow of contrast material from the renal cortex into the medulla with subsequent homogeneous enhancement of the whole-kidney parenchyma. The parenchymal phase began with a minimal signal drop at the end of the initial vascular peak giving a peak–trough pattern (Fig. 2A). This was followed by a subsequent slower linear increase up to a second peak, which was seen at 162.6 ± 9.2 sec. The peak–trough pattern was observed in 89.5% (17/19) of the studies. This was present in all patients who had an injection administered with an injector (n = 11) and was absent in the two studies performed with manual injection of contrast material. Finally, contrast material was excreted into the collecting system, with bright signal seen first in the renal calices, then the renal pelvis, subsequently in the ureter, and finally the bladder. The excretory phase was characterized by a prompt concave decline in the third segment of the renographic curve, which was less prominent than that seen in the corresponding radionuclide study.
|
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
It is only recently that Rohrschneider et al. [10] studied 62 pediatric patients using a 0.5- T scanner and a full dose of gadopentetate dimeglumine (0.1 mmol/kg of body weight). These researchers found that MR imaging could reveal split renal function and assess excretory status in a way similar to that of radionuclide studies. Further, these researchers found that MR imaging can show kidney and urinary tract morphology. An earlier study of 29 adult patients by Taylor et al. [12] using a 1-T MR scanner has also shown good correlation in the split renal function between the two techniques. They reported that correlation was best when a low dose of gadopentetate dimeglumine was used. However, their technique was compromised by suboptimal temporal resolution. They used time intervals of 30 sec between consecutive image acquisitions, in which the true peak signal intensity may be missed, resulting in erroneous split function calculation.
The results of our study show that MR renography using dynamic gradient-echo sequences with high flip angles and low-dose gadopentetate dimeglumine performed with a 1.5-T MR scanner is a feasible alternative to radionuclide renography. We could obtain signal intensity–time renographic curves similar to those of radionuclide renography (Fig. 2A, 2B).
Principle of MR Renography
MR renography is based on the principle that the intensity of MR signal
generated is directly proportional to the concentration of gadopentetate
dimeglumine in the kidney. The signal intensity–time curve hence
obtained will be a reflection of the renal function and excretion. Unlike
radionuclide or studies using iodinated contrast material in which the
brightness or intensity of the radiotracer or contrast material is directly
proportional to the concentration of the radiotracer or contrast material
used, in MR imaging the signal intensity obtained is a measurement of the
interactions of T1 and T2/T2* effect in the ROI.
An MR imaging protocol in which the signal intensity acquired is linear
with gadopentetate dimeglumine concentration should be used. Gadopentetate
dimeglumine has both a T1- and T2-shortening paramagnetic effect. T1
shortening is desired because it enhances the signal intensity of urine. T2
shortening destroys the positive contrast material effect of urine and results
in a signal void. Hence, scan parameters should enhance T1 but reduce T2/T2*
effects. This may be achieved by using a short-TE and high flip angle
[23]. Short-TE reduces the T2
effect and makes MR signal predominantly T1-weighted
[12,
22]. High flip angle (
40°) potentiates the T1-shortening effect of a short-TE and reduces the
scan sensitivity to T2* effect
[24]. It also reduces the
T2-shortening effect of gadopentetate dimeglumine at a high concentration
[11]. We used a short-TE time
of 4.2 mm/sec and adopted a flip angle of 60° in our study. We did not use
a flip angle of 90° because this has been found to result in slice profile
distortion [12,
25].
Furosemide is a loop diuretic that blocks the Na+/K+/2Cl-symporter in the ascending limb of the loop of Henle leading to rapid retention of water inside the tubules. Urine volume and flow are markedly increased resulting in dilution of the excreted gadopentetate dimeglumine concentration. Furosemide also causes a rapid and uniform distribution of gadopentetate dimeglumine inside the urinary tract. These effects optimize the T1-enhancing properties of gadopentetate dimeglumine and reduce the T2* effect in the gadopentetate dimeglumine–enhanced urine [26]. Furosemide is administered 15 min before gadopentetate dimeglumine injection so that its effect will peak at the time when gadopentetate dimeglumine is injected.
We found in our volunteers, particularly in the elderly, that at full dose (40 mg) of furosemide administration, patient discomfort tended to be worse due to an overdistended bladder resulting in failure to maintain a full bladder until completion of the study. This discomfort has also been observed by other researchers who reported that furosemide stimulation is a high-risk factor for bladder overfilling causing inhibition of urinary drainage [10]. It has also been reported that visualization of the distal ureter and bladder is variably impaired by artifacts caused by turbulence in the region from the furosemide-induced rapid diuresis [27]. Moreover, previous researchers have found that a positive interaction between furosemide and gadolinium already occurs at low IV diuretic doses [26]. For these reasons, we decided to perform our study using a low-dose IV furosemide of 0.3mg/kg of body weight to a maximum of 20 mg.
We used a low dose of gadopentetate dimeglumine (0.025 mmol/kg–1) for our study. The signal intensity–time curve obtained with low-dose gadopentetate dimeglumine has been shown to correlate best with the renographic curve obtained with radionuclide renography. In addition, the T2* effect is not significant with this concentration range [12]. High gadopentetate dimeglumine concentration is undesirable because this causes T2 shortening and magnetic susceptibility artifacts, which further contribute to the T2 effect [4, 28].
Split Renal Function
Split renal function is calculated from areas under the curve taken at
1–2 min after injection of gadopentetate dimeglumine. Previous
researchers [12] have reported
that at the injected dose of 0.025 mmol/kg–1, the
concentration of gadopentetate dimeglumine is likely to remain low during this
time and hence is linearly related to the signal intensity.
Split renal function determined from the MR intensity–time curve gives the mean signal intensity per voxel of tissue, compared to radionuclide renography, which computes the sum of all counts in the whole kidney [12]. This determination enables evaluation of per unit volume of regional function, which is not possible with radionuclide renography. The overall volume-corrected split renal function ratio is dependent on both per unit voxel function and total volume of the kidney. Methods that determine one of these components without taking into consideration the other factor will result in erroneous split renal function values.
The value obtained should be generated from an ROI drawn over the renal parenchyma. In patients with significant hydronephrosis, the distended non–contrast material–filled renal pelvis will contribute a large area of low signal to the ROI interpreting if the ROI is taken over the entire kidney including the cortex and renal pelvis. This reading may result in an erroneous low value for the split renal function. Our study revealed that in hydronephrotic kidneys with thinned cortexes, split renal functions between the two techniques may differ by approximately 10%. This may be attributed to the capacious renal pelvis, which may attenuate radiotracer intensity in radionuclide renographic studies resulting in an erroneous value. Further, a small inaccuracy in the calculation of the ROI area in a small kidney will contribute to a significant percentage of difference in volume measurement, affecting the computation of overall split renal function between the two kidneys. The discrepancy in the split renal function in hydronephrotic systems has also been reported by Rohrschneider et al. [10].
Volume Calculation
The overall split renal function is volume-dependent (renal function = area
under the curve x total volume of kidney). An inaccurate estimation of
renal volume will result in an erroneous measurement of split renal function.
Measurements using bipolar length and area in the parenchymal rim in the mid
coronal plane are not directly proportional to the renal volume because the
kidneys are three-dimensional organs and asymmetric bilaterally. The kidney's
transverse width also contributes significantly to the total renal volume. MR
imaging allows the use of the voxel count and slice summation method
[14], which we also used. The
advantage of using this method is that the shape of the kidney is irrelevant.
The drawback of this method is that it has to be done manually and hence is
operator-dependent and time-consuming. Further, in cases in which there are
multiple space-occupying lesions (e.g., polycystic kidneys), accurate
calculation of volume may be difficult manually. The availability in the near
future of automated or semiautomated software for volume calculation would
greatly improve the efficacy and accuracy of renal volume calculation.
Another potential source of inaccuracy that may result in erroneous split
renal function calculation is the time interval between consecutive MR images.
A short time interval between two consecutive MR renograms is desired because
a long scan time between consecutive images will result in a loss of critical
information. The true peak intensity of the second segment may be missed,
resulting in a low renal function being obtained incorrectly. Magnetic
susceptibility from medical devices in or around the kidneys, such as a
double- stent as in our case, is another potential source of
error.
Kikinis et al. [8] found superior image quality in patients who were in respiratory apnea compared with patients with normal breathing during image acquisition. This breath-hold technique would be difficult to achieve without compromising temporal resolution needed for accurate split renal function calculation. We allow our patients shallow and continuous breathing during acquisition of the dynamic MR renographic sequences. The relatively low respiratory motion of the kidneys in the confines of the retroperitoneal space allows the generation of an MR signal intensity–time curve satisfactory for determination of the split renal function. Non–breath-hold MR renographic techniques were also adopted by Rohrschneider et al. [10] and Taylor et al. [12] and were shown to have good results comparable to radionuclide renography.
Renographic Curve
We observed several differences between the two techniques. The
time-to-peak of the parenchymal transit as reflected by the second segment
occurs earlier in MR renography compared to radionuclide renography. This is
noted within 3 min in our studies. The normal time-to-peak is about 3–5
min for radionuclide renography
[1,
29]. This observation was also
noted by Taylor et al. [12],
who observed a time-to-peak of approximately 3 min when MR renography was
performed at a concentration of 0.025 mmol/kg of body weight. They attributed
this to the different injection procedures used in the two techniques. In MR
renography, a large volume of contrast material is delivered through a long
tubing, whereas in radionuclide renography, a smaller injection is made
directly into the vein.
A peak–trough pattern is noted at the commencement of the parenchymal phase when a good bolus is achieved (Fig. 2B). A similar dip in signal intensity has also been reported by other studies [4, 8, 9, 21]. This does not occur in radionuclide renography because of T2 shortening and the magnetic susceptibility effect with the arrival of high concentrations of gadopentetate dimeglumine in the renal cortex [4, 10, 30]. Ros et al. [20] reported that with a poorly injected IV bolus, distinct enhancement peaks were often obscured. The presence of a peak–trough pattern in MR renography may be used as a good artifact to serve as a marker of good bolus injection. This phenomenon appears to be MR-sequence dependent. Rohrschneider et al. [10, 11] and Pettigrew et al. [31], all of whom performed MR renography with a magnetic field strength of 0.5-T, obtained similar findings with gradient-echo sequences but not with a dynamic T1-weighted two-dimensional fast field-echo technique.
A higher baseline is noted in the excretory curve obtained with MR renography in our study. Rohrschneider et al. [10] and Taylor et al. [12] described a similar trend. This may have been caused by the different agents used for dynamic MR renography (gadopentetate dimeglumine) and radionuclide renogram (99mTc mercaptoacetyltriglycine). Gadopentetate dimeglumine is eliminated by glomerular filtration [32], whereas 99mTc mercaptoacetyltriglycine is predominantly excreted by tubular secretion and has a more efficient extraction [33].
Limitations of our MR imaging scanner and its operating software allow acquisition of only 128 consecutive images. Ten-second intervals between images permit a total scan time of only approximately 20 min. This is shorter than the 25 min used for radionuclide renography, resulting in an incomplete acquisition that misses the concluding segment of the excretory curve. Another reason for the inconsistencies could be that MR renography scans only a single slice through the renal hilum, whereas radionuclide renography obtains its total radiotracer count from the whole kidney. MR renography may have excluded the parenchyma in the regions that tend to enhance faster than the more central regions of the kidney, resulting in reduction in peak intensity and hence underestimation of the gadopentetate dimeglumine excretion [12].
Dynamic Serial Images
The MR renographic serial images were superior in the depiction of renal
morphology compared with radionuclide renograms. Renal scarring and cortical
thickness were well seen (Fig.
5A,
5B,
5C). They also allowed
visualization of background anatomy.
|
|
|
We observed dynamic changes, similar to those obtained on CT with contrast enhancement, showing progressing enhancement of the renal parenchyma from the cortex to the medulla and finally the renal pelvis as contrast material passed through the kidney. These are in contrast to methods that used a combination of gradient-echo technique and high gadopentetate dimeglumine concentration (0.1 mmol/kg of body weight), in which a peripheral dark band that moved centripetally from the cortex to the medulla is observed with the dynamic images [4, 5, 21]. This dark band appearance is concentration-dependent and is not seen with 0.01 mmol/kg of body weight of gadopentetate dimeglumine [5]. It is also MR sequence–dependent; it is not seen with fast spin-echo sequences despite administration of a high dose of gadopentetate dimeglumine (0.1 mmol/kg of body weight) [10, 31].
The short-acquisition interval (10 sec) between consecutive images used by our technique permits near real-time evaluation of the urinary system and visualization of contrast material passage through the kidneys. This is not possible with other imaging techniques such as CT or excretory urography without exposing the patients to excessive amounts of ionizing radiation. We observed a whirlpool phenomenon that, to our knowledge, has not been described (Fig. 6). The patient with this phenomenon had an obstruction of the ureteropelvic junction. In patients with significant ureteropelvic junction obstruction, the urine will encounter a focus of high resistance at the junction of the renal pelvis and the ureter. Its forward flow is blocked as the "gate" into the ureter is closed. The urine swirls in a circular motion as it is forced to return via the path of least resistance into the renal pelvis. We suspect that this observation occurs in situations in which there is severe obstruction at the ureteropelvic junction. For patients in whom the whirlpool phenomenon is noted, a more aggressive treatment may be advised.
|
Further, analysis of the serial renographic images is useful in patients who have poor renal function to ensure that an obstructive excretory pattern is not due to a poor diuretic response. Assessment of the dynamic images also helps to establish that an abnormal or equivocal excretory curve pattern is not the result of markedly rigorous respiration causing excessive renal movement. Also in patients with ureteropelvic junction obstruction, review of the dynamic images helps to determine that the ROI is correctly drawn to include the renal pelvis up to the level of obstruction [34, 35].
Limitation and Future Direction
An obstacle to the routine use of MR imaging for renal function evaluation
is the cost of the examination. More recently, an MR imaging technique using
low-dose gadopentetate dimeglumine (0.06 mmol/kg of body weight) MR urography
after MR renography with subsequent additional treble-dose gadopentetate
dimeglumine (0.3 mmol/kg of body weight) for MR angiography has been reported
[23]
(Fig. 7). This allows
functional, anatomic, and vascular assessment of the urinary system. Studies
by Katzberg et al. [6] and
Niendorf et al. [7] have shown
that single-kidney filtration fraction and glomerular filtration rate
determination using T1 measurements of flowing blood in the renal vein and a
systemic vessel are feasible. The potential of a comprehensive all-in-one
assessment of the urinary system by MR imaging may in the future be more
cost-effective and convenient for the patient, who might otherwise undergo
multiple examinations for a complete evaluation of the renal system. The cost
of a single MR imaging examination is likely to be economical compared with
the overall cost of all the different imaging techniques.
|
Some patients are claustrophobic. This procedure also requires patient cooperation. Tachypneic and restless patients are poor candidates for the procedure.
The need for high temporal resolution places limitations on the MR
renographic technique on the basis of the present available hardware and
software technology. The required temporal resolution implies that only
limited representative slices through the kidney instead of its whole volume
can be acquired. Therefore, an inaccurate conclusion will be obtained for
patients with regional abnormalities. The study would be improved by
techniques (including two-dimensional or three-dimensional multislice methods)
that can acquire repetitive series of multislice images to give complete
coverage of the kidney without compromising signal-to-noise ratio in the
images in a short temporal resolution (
10 sec) over a sufficient time
span ( 25 min) so as to produce a renographic curve that is similar to
that in a radionuclide study.
An accurate renal-volume calculation is also paramount to the success of the MR renographic technique. This remains a technical challenge and is a significant issue in patients who have multiple renal lesions. The postprocessing of data at this point in time remains tedious. However, with advancement in the hardware and software technology, we expect the procedure to improve.
In conclusion, our study shows that MR renography is a potential feasible alternative to radionuclide renography, showing good correlation and agreement between the two techniques. Our technique uses dynamic MR gradient-echo sequences with low-dose IV furosemide and gadopentetate dimeglumine (0.025 mmol/kg of body weight) and can produce an intensity–time curve and serial dynamic images of the urinary system, which allow assessment of split renal function and urinary excretory status, in a way similar to that of radionuclide renography. Our study was performed using an MR scanner that is commonly available.
A good bolus injection, temporal resolution, and accurate estimation of the renal parenchymal volume are essential for an accurate calculation of the split renal function. Advancement in imaging technology that allows even better temporal resolution and complete renal-volume coverage will make this technique more robust.
The serial images obtained with MR imaging give a dynamic renal enhancement pattern that is similar to that of CT nephrography. The morphology of the kidneys and the urinary outflow tract is far superior to that seen on radionuclide renography. MR renography also shows background anatomy and does not require ionizing radiation.
In addition, our technique allows for the easy addition of MR urography and MR angiography during the same examination. The potential of an all-in-one investigative tool that will allow a comprehensive vascular, anatomic, and functional assessment without ionizing radiation makes MR renography vital for urologic imaging. Its potential is enormous.
Acknowledgments
We thank Fergus V. Coakley, associate professor and chief, Abdominal
Imaging Section, Department of Radiology, University of California, San
Francisco Medical Center, for his advice in the preparation of this
manuscript.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  P.-H. Vivier, M. Dolores, I. Gardin, P. Zhang, C. Petitjean, and J.-N. Dacher In Vitro Assessment of a 3D Segmentation Algorithm Based on the Belief Functions Theory in Calculating Renal Volumes by MRI Am. J. Roentgenol., September 1, 2008; 191(3): W127 - W134. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Boss, P. Martirosian, M. Gehrmann, F. Artunc, T. Risler, N. Oesingmann, C. D. Claussen, F. Schick, K. Kuper, and H.-P. Schlemmer Quantitative Assessment of Glomerular Filtration Rate with MR Gadolinium Slope Clearance Measurements: A Phase I Trial Radiology, March 1, 2007; 242(3): 783 - 790. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. V. Prasad Functional MRI of the kidney: tools for translational studies of pathophysiology of renal disease Am J Physiol Renal Physiol, May 1, 2006; 290(5): F958 - F974. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
= right kidney, 
= left kidney.
