| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Diagnostic Radiology, Justus-Liebig-University, Klinikstrasse
36, 35392 Giessen, Hessen, Germany.
2 Mathematical Institute, Justus-Liebig-University Giessen, Giessen,
Germany.
Received August 31, 2006;
accepted after revision November 21, 2006.
Address correspondence to C. Schneider
(christian.schneider{at}radiol.med.uni-giessen.de).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. We examined 12 healthy adult volunteers. Blood flow in one renal artery was measured using cine phase-contrast imaging (1.5-T MR system). Each volunteer received 120 mL of isotonic sodium chloride on study day 1 and 120 mL of a low-osmolar, nonionic, iodinated contrast medium (iomeprol, 400 mg I/mL) on study day 2. Repetitive measurements were performed before (up to five measurements in 5 minutes) and after (up to 13 measurements in 30 minutes) the injection was started.
RESULTS. Mean basal renal artery blood flow was 664 mL/min. In response to the injection of the test substances, we found a significantly larger decrease in average renal blood flow for contrast medium than for sodium chloride (31.9 mL/min vs 18.3 mL/min, p = 0.0481). Furthermore, in analyzing the measurements at early time points, we found a significant decrease (11.4% ± 4.7% [SD]; Bonferroni-corrected, p < 0.05) in renal blood flow 2 minutes after the injection of the contrast medium was started. Sodium chloride did not produce a significant effect at any time.
CONCLUSION. Cine phase-contrast MRI can measure a decrease in renal blood flow in humans in response to an IV injection of iodinated radiographic contrast medium. Therefore, cine phase-contrast MRI can be a helpful and noninvasive tool for further investigations of contrast media-induced changes in human renal blood flow and their possible impact on the development of contrast-induced nephropathy.
Keywords: contrast-induced nephropathy • contrast media • hemodynamics • kidney • MR technique • renal disease
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Compared with the animal experiments mentioned previously, in which small amounts of high-osmolar contrast media were injected directly into one renal artery, the clinical use of contrast media in humans is actually quite different. First, low-osmolar nonionic contrast media have replaced high-osmolar contrast media. Second, a difference between the hemodynamic reactions after contrast medium injection in animals, such as dogs or swine, and in humans can be assumed. Third, the injection protocol—namely, the amount of contrast medium and the site of injection—is different in most cases because larger volumes of contrast medium are usually injected either IV (e.g., in CT) or intraarterially (e.g., in coronary angiography).
In the past, noninvasive measurement of RBF using cine phase-contrast MRI (PC-MRI) was developed and validated [17, 18]. The aim of the present study was to investigate the effect of IV-injected, low-osmolar nonionic radiographic contrast media on human RBF using a noninvasive PC-MRI technique.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Twelve adult subjects were included (four women, eight men). The age range was 23-28 years (mean age, 25 years). Mean body weight was 71 kg (range, 55-84 kg) and mean body height was 179 cm (range, 165-189 cm). Examinations started in October 2004 and lasted until January 2005.
Study Design
Each volunteer was examined on two study days that were at least 1 week
apart. On study day 1, the volunteer received an IV injection of 120 mL of
isotonic sodium chloride solution (NaCl 0.9%, B. Braun Melsungen AG;
osmolarity, 309 mOsm/L) at a flow rate of 3 mL/s. On study day 2, the
volunteer received 120 mL of iomeprol 400 (Imeron 400, Altana Pharma AG;
osmolarity, 726 mOsm/L; 400 mg I/mL) at the same flow rate. Iomeprol 400 is
routinely used in our department, and 120 mL represents a mean dose for
whole-body CT scans. The MR examinations took place between 7:00 and 9:00 am.
The volunteers had all fasted overnight, and neither caffeinated nor
nicotine-containing or other stimulating products were allowed before the
examination. The volunteers were instructed to drink 500 mL of fluid such as
water or juice at least 1 hour before the examination to provide constant
conditions of hydration and renal activity.
The PC-MRI measurements were performed five times before the injection started, each 1 minute apart, at time points t = -1 to -5, referred to as basal RBF measurements. The first measurement after injection of NaCl or contrast medium (t = 1) was performed 10 seconds after the injection was started. Subsequent measurements were started at the beginning of the second, third, fourth, and fifth minutes (t = 2-5) and then every 3 minutes up to 30 minutes after injection (t= 8-30). Each measurement lasted approximately 17 seconds. After the first five experiments, we extended our measurement schedule to increase the validity of the statistical analysis. Therefore, a few RBF measurements are missing in three experiments investigating NaCl and in two experiments investigating contrast medium (see also Table 1). Altogether, for each subject we obtained two sequences of (up to) 18 RBF measurements for each study day, resulting in a total of (up to) 36 RBF measurements for each subject.
|
Measurement of RBF using MRI
MR measurements were performed on a whole-body scanner at 1.5 T (Gyroscan
Intera, Philips Medical Systems). Each volunteer was placed supine with the
arms alongside the body. An 18- or 20-gauge needle was placed in the cubital
vein. The leads for a three-channel ECG were placed. A multichannel
phased-array coil (SENSE-cardiac coil, Philips) with two separate coils was
placed ventrally and dorsally to the kidneys. After scout images were obtained
in coronal, transverse, and sagittal directions for localization of the renal
arteries, a multishot gradient-echo sequence with magnetization preparation
using prepulses was used to obtain the required plane for the phase-contrast
measurements (balanced Turbo Field Echo sequence (bTFE) [Philips]; slice
thickness, 3 mm; 30 slices; matrix, 320 x 320; field of view, 400 mm;
flip angle, 90°; scanning duration, 3 minutes 48 seconds using respiratory
gating; TR/TE, 5.8/2.3). The bTFE sequence was first performed in the axial
direction. Then oblique coronal images parallel to the course of the right
renal artery were acquired. Two volunteers had an accessory right renal
artery. In these cases, the larger-caliber artery was chosen. In one
volunteer, the left renal artery had to be measured because of a tortuous
right renal artery. Finally, for measuring RBF, the phase-contrast scan was
placed perpendicular to the course of the renal artery as shown in the oblique
coronal bTFE sequence.
We used a cine phase-contrast MR velocity mapping technique. A 2D segmented gradient-echo sequence was used consisting of eight segments (shots), each consisting of 12 phase-encoding steps (Turbo Field Echo sequence [TFE], Philips). We acquired phase-difference images spanning the whole cardiac cycle. Twenty frames were generated per cardiac cycle with cardiac synchronization provided by the ECG. Magnitude and phase-difference images were reconstructed for each frame. The duration of one measurement was approximately 17 seconds, depending on the individual cardiac frequency. The measurements were performed using a deep inspiration breath-hold technique because it has been shown that measurements during apnea are more accurate than measurements during continuous breathing [19]. Phase velocity mapping was performed with a TR/TE of 5/2.8 and a flip angle of 15°. Slice thickness was 8 mm and field of view, 350 mm. Measured voxel size was 1.99 x 2.57 x 8.00 mm, and reconstructed voxel size, 1.37 x 1.37 x 8.00 mm. These parameters were chosen carefully to minimize systematic and random measurement error because an adequate spatial resolution is crucial to minimize systematic and random error of flow measurements [20]. Velocity encoding was set to 150 cm/s because normal peak velocity in the renal arteries often exceeds 100 cm/s [17]. Hence, no aliasing was observed.
The velocity in each voxel was obtained from the measured phase difference. For each cine frame, a region of interest (ROI) was placed over the renal artery using the modulus image. Initially, the ROI was placed manually in one frame, making sure that it enclosed the vessel totally but remained as small as possible to minimize measurement errors [20]. After placing the first ROI in one frame, a semiautomated method provided by the software of the MR workstation (EasyVision, release 5.1, Philips) was used to copy this ROI to all other 19 frames. All ROIs had the same size and were inspected for proper positioning. In many examinations, a systematic movement of the renal artery was seen after the cardiac cycle. In cases of relevant movement, the ROI was adjusted to optimally enclose the renal artery. After the ROIs were placed, the software generated a value for mean flow in milliliters per second.
Statistics
Our RBF data exhibited a hierarchic correlation structure, with two levels
in each individual: On the first level, correlations exist among measurements
in each sequence, and on the second level, between two sequences (during
different treatments) taken in the same individual. Missing values were
considered as missing completely at random.
For the statistical analysis of the RBF time courses, we set up a regression equation to model RBF trends across time and differences due to treatments (NaCl or contrast medium). In particular, a two-level linear mixed-effects regression model was used [21]. The model included the following effects: a treatment-dependent, piecewise linear function of time, modeled by fixed effects allowing for a "jump" in RBF immediately after injection at t= 0 followed by a linear time trend. To account for the interindividual variation among the average basal RBF measurements, a random vertical shift was included; and for the intraindividual between-treatment variation among the RBF measurements, a random treatment-by-time interaction (also piecewise linear) was included. Statistical analyses were performed using the software R (R Development Core Team) [22], which includes the package nlme [23].
Using this model, the following three questions were investigated: Is there a change in average RBF due to injection of NaCl? Is there a change in average RBF due to injection of contrast medium? Is there a difference between the two treatments with respect to the average RBF time course after injection?
In addition to these investigations, a separate analysis of only the five early RBF measurements at 1, 2, 3, 4, and 5 minutes after injection was conducted to discover which of the early mean relative changes in RBF differ between NaCl and contrast medium. Every mean relative change was calculated from 10 to 12 individual relative changes, each from one individual. The individual relative change in RBF at time t in each treatment was defined by [(RBF at time t) - (individual average basal RBF)]/(individual average basal RBF). The individual average basal RBF was calculated as the mean of the five measurements before injection of either NaCl or contrast medium and was different for each individual and treatment. To account for the correlation among measurements in the same sequence, the five early relative changes in RBF were treated as a vector; and for the comparison of the two treatments, we considered the differences of the vectors of early relative changes in RBF between contrast medium and NaCl for every individual.
Here, we addressed the following three questions: At which time during the first 5 minutes is the relative change in RBF: during treatment with NaCl significantly less than zero? during treatment with contrast medium significantly less than zero? and during treatment with contrast medium significantly greater than during treatment with NaCl?
Identification of those time points was done using the Bonferroni method for simultaneous upper 95% confidence limits. These limits were determined in order to be able to make statements such as, "We are 95% confident that the RBF is changed during treatment with NaCl at minute x on average by at most y%."
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Decomposition of the overall variability of the total of 116 basal RBF measurements into three variance components gave the following estimates for the respective SDs: interindividual SD of average basal RBF was 104 mL/min (resulting in a relative SD of 15.7% of the overall mean basal RBF); intraindividual between-treatment SD of average basal RBF was 62 mL/min (relative SD: 9.3% of the overall mean basal RBF); and intraindividual, within-treatment SD of basal RBF measurements was 31 mL/min (relative SD: 4.7% of the overall mean basal RBF).
This is significantly different (p = 0.0003) from the SD of 60 mL/min of the RBF measurements after injection (relative SD: 9% of the overall mean basal RBF). This means that after correcting for time trends and treatment differences, the remaining variability of RBF after injection is significantly larger than before. No significant difference was seen between the two treatments with respect to the SD after injection (p = 0.4931).
With regard to our first three questions to be investigated we found:
First, a nonsignificant and numerically weak decrease of 18.3 mL/min (p = 0.0870) in the average RBF after injection of NaCl, followed by a compensating rise of 1.3 mL/min (p = 0.0275), so that average RBF is back to the original basal level after approximately 15 minutes.
Second, a significant decrease by 31.9 mL/min (p = 0.0031) in average RBF after injection of contrast medium without a return to the original average level during the observed 30 minutes (slope of average RBF after injection, -0.3 mL/min; p = 0.5788).
Third, a significant difference in average postinjection RBF time courses between treatments (p = 0.0481).
As an example, Figure 1A shows the fitted regression curves for one individual, and Figure 1B presents the relative changes in RBF from study day 1 (injection of NaCl) and study day 2 (injection of contrast medium) for the same individual. Figure 2A, 2B shows for each study day the averaged results of all 12 individuals for the time points considered. Table 1 contains the numeric results of both study days, showing the mean relative changes in RBF (relative to individual average basal RBF) at every time point considered and their respective SDs. In both groups, the SDs of each time point appear visibly larger after injection than those of the basal measurements, as already mentioned.
|
|
|
|
The average relative change in RBF during treatment with contrast medium found at 2 minutes after injection was a decrease of 11.4%. Later measurements (i.e., for t >5) show average relative changes in RBF after contrast medium injection between -4.2% and -8.6%.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
RBF can be reliably measured by means of cine PC-MRI [17, 18]. When compared with measurements of p-aminohippurate clearance as an established method for the determination of RBF, the PC-MRI results showed a good correlation (r = 0.93) in human experiments [17]. In animal experiments, PC-MRI measurements of RBF showed an excellent correlation (r = 0.99) compared with invasive measurements using a transit-time ultrasound flow meter [18]. The accuracy of PC-MRI measurements of RBF depends on different factors, mainly concerning spatial resolution. Thus, an adequate imaging strategy is crucial. In the present setting, an error of less than 5% can be assumed [20]. Furthermore, the proper positioning of the measuring plane as described in the Subjects and Methods section plays an important role for the accuracy and precision of flow measurements because an offset angle of 11° may cause an error of up to 3% [20]. The measuring plane is therefore specified by the course of the vessel. This implies that simultaneous measurements in a reference vessel were not possible.
A few studies have shown an effect of iodinated contrast media on proton relaxation in MRI with a relative reduction of T1 and T2 relaxation times [24, 25]. This effect was concentration-dependent and diminished as the concentration of contrast medium decreased. Relevant effects on T1 and T2 relaxation times would result in changes of the amplitude of the blood signal in the vessel, but we did not observe any changes of the signal amplitudes at all when performing comparative measurements before and after the administration of contrast medium or NaCl. It thus appears unlikely that PC-MRI is affected by iodinated contrast media in the present setting. Finally, after theoretic considerations, if any effect existed, it would result in falsely increased values for velocity and RBF. Thus, the observed effect on RBF in the present study cannot be caused by but may rather be diminished by iodinated contrast medium (Kooijman H, personal communication).
In 1968, Talner and Davidson [6] reported on experiments measuring the effect on RBF after contrast medium injection into one renal artery of dogs [6]. The test substances were injected as a bolus of 0.3 mL/kg of body weight in 2-2.5 seconds. As test substances, high-osmolar ionic contrast media (meglumine diatrizoate 50%, meglumine iothalamate 60%, and sodium diatrizoate 60%) and hypertonic NaCl 4.5% solution were used. Each of the four agents had an osmolarity of approximately 1,400 mOsm/L [6]. RBF was measured 1, 3, 7, and 13 minutes after injection using a dye dilution technique. As a result, RBF showed a statistically significant decrease—between 10% and 20%—in all four substances in the first minute, followed by different changes in the subsequent measurements that were not statistically significant. In 1970, Caldicott et al. [1] reported a similar experiment. Using the same contrast media as Talner and Davidson [6], Caldicott et al. measured RBF continuously by means of an electromagnetic flow meter (diameter, 3 mm) that was placed inside the renal artery. They rapidly injected a 3-mL contrast medium bolus. Having a better time resolution than Talner and Davidson, they found a biphasic reaction of RBF in response to the contrast medium injection, with an initial increase (36.0% ± 8.1%) that lasted for less than 30 seconds, followed immediately by a decrease (27.6% ± 4.0%) at an average of 45 seconds after injection and a second, smaller flow decrease of less than 10% between 2 and 15 minutes after injection [1]. Injection of hypertonic sodium chloride (NaCl 4.5%) produced a less pronounced but qualitatively identical reaction.
The maximal concentration of contrast medium in blood after IV injection occurs a few seconds after the end of the injection, independent of the duration of injection [26]. In the two animal experiments mentioned previously, the injection lasted only a few seconds, so the maximum blood concentration can be expected at around t = 0 minutes. Allowing for the course of RBF found by Caldicott et al. [1], the following points should be considered: Because of the low time resolution in the present study, with measurements lasting approximately 17 seconds every minute, the rapidly passing increase in RBF immediately after injection could not be detected. An explanation for the observed relatively slight, but yet significant maximal decrease in RBF of merely 11.4% in the second minute might be that this measurement was performed in part during the equalization phase of the initial increase and the subsequent decrease in RBF. Therefore, again due to the relatively long measurement duration, the transient maximal decrease in RBF could not be measured exactly to its whole magnitude. Altogether, after this argumentation, we assume that our results are in good correlation with the experimental studies by Talner and Davidson [6] and Caldicott et al. [1], and that failure to exactly reproduce especially the results of Caldicott et al. is due to the relatively low time resolution of PC-MRI.
In contrast to the results of Talner and Davidson [6] and Caldicott et al. [1], we observed a prolonged decrease in RBF after injection of contrast medium that was not yet compensated at the end of the observation period. This finding might be explained by the larger volume of contrast medium used in the present study and the different administration route, both of which result in a different distribution throughout the body and a prolonged elimination of contrast medium [26].
As reviewed in depth elsewhere, the underlying pathophysiologic mechanisms that cause contrast-induced nephropathy remain unknown [27]. Because the contribution of changes in RBF to the development of contrast-induced nephropathy is still under discussion, further investigations on that issue are needed to understand the nature of contrast-induced nephropathy and to develop further strategies to prevent it. Our findings cannot answer the question of whether the relatively slight and transient changes in RBF contribute to contrast-induced nephropathy at all because we did not perform follow-up examinations (e.g., monitoring of serum creatinine) and because we did not investigate individuals with a preexisting impairment in renal function, who actually are known to be more susceptible to contrast-induced nephropathy [28]. In fact, our results clearly indicate that PC-MRI is capable of measuring an effect on RBF in response to an IV injection of contrast medium in healthy young people. PC-MRI can thus be a helpful and noninvasive tool for further investigation of contrast medium-induced changes in human RBF and their possible impact on the development of contrast-induced nephropathy. Additional studies are now needed to verify whether the changes in RBF are reproducible, and maybe even more pronounced, in individuals at risk for contrast-induced nephropathy with preexisting renal dysfunction.
PC-MRI is basically suitable for all patients who have no contraindications to MRI and who are able to lie supine and hold their breath for a short period. Limitations of PC-MRI are a relatively low time resolution, so transient changes may be missed, and a relatively long measurement duration, which to a certain extent causes artifacts due to respiratory movement. Both limitations may be remedied in part with further improvements in MRI techniques in the future.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  R. C. Harmon, S. P. Duffy, M. V. Terneus, J. G. Ball, and M. A. Valentovic Characterization of a novel model for investigation of radiocontrast nephrotoxicity Nephrol. Dial. Transplant., March 1, 2009; 24(3): 763 - 768. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
])
and 497 ± 14 mL/min on study day 2 (contrast medium measurements
[
]).
