| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Genitourinary Imaging |
1 From the Departments of Radiology, Neuroradiology, and Nuclear Medicine (H.C.T., S.S., P. Vock); Clinical Research (D.Z., C.B., P. Vermathen); and Nephrology and Hypertension (U.E., L.H., F.J.F.), University Hospital of Bern, Inselspital, Freiburgstrasse 10, CH-3010 Bern, Switzerland; and Department of Radiology, University Hospitals Leuven, Leuven, Belgium (F.D.K.). Received January 17, 2006; revision requested March 22; revision received April 8; final version accepted June 8. H.C.T. supported by a research grant from the Swiss National Foundation, National Center of Competence in Research, "Computer-aided and image-guided medical interventions", NCCR CO-ME. F.J.F. supported by grant no. 31-102153 from the Swiss National Foundation for Scientific Research. P. Vermathen and H.C.T. supported by grant no. 320000-111959/1 from the Swiss National Foundation for Scientific Research. Address correspondence to H.C.T. (e-mail: Harriet.Thoeny{at}insel.ch).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
Materials and Methods: The local ethics committee approved the study protocol; patients provided written informed consent. Fifteen patients with a renal allograft and in stable condition (nine men, six women; age range, 20–67 years) and 15 age- and sex-matched healthy volunteers underwent DW and BOLD MR imaging. Seven patients with renal allografts were examined twice to assess reproducibility of results. DW MR imaging yielded a total apparent diffusion coefficient including diffusion and microperfusion (ADCtot), as well as an ADC reflecting predominantly pure diffusion (ADCD) and the perfusion fraction. R2* of BOLD MR imaging enabled the estimation of renal oxygenation. Statistical analysis was performed, and analysis of variance was used for repeated measurements. Coefficients of variation between and within subjects were calculated to assess reproducibility.
Results: In patients, ADCtot, ADCD, and perfusion fraction were similar in the cortex and medulla. In volunteers, values in the medulla were similar to those in the cortex and medulla of patients; however, values in the cortex were higher than those in the medulla (P < .05). Medullary R2* was higher than cortical R2* in patients (12.9 sec–1 ± 2.1 [standard deviation] vs 11.0 sec–1 ± 0.6, P < .007) and volunteers (15.3 sec–1 ± 1.1 vs 11.5 sec–1 ± 0.5, P < .0001). However, medullary R2* was lower in patients than in volunteers (P < .004). Increased medullary R2* was paralleled by decreased diffusion in patients with allografts. A low coefficient of variation in the cortex and medulla within subjects was obtained for ADCtot, ADCD, and R2* (<5.2%), while coefficient of variation within subjects was higher for perfusion fraction (medulla, 15.1%; cortex, 8.6%). Diffusion and perfusion indexes correlated significantly with serum creatinine concentrations.
Conclusion: DW and BOLD MR imaging are feasible and reproducible in patients with renal allografts.
© RSNA, 2006
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
Diffusion-weighted (DW) MR imaging yields the apparent diffusion coefficient (ADC) as a quantitative parameter, which reflects the microenvironment of diffusing water molecules. In addition to diffusion information, DW MR imaging simultaneously provides information on microcirculation—including capillary perfusion (5–7)—provided diffusion and microcirculation can be separated.
Since the main kidney functions are related to transportation of water (glomerular filtration, active and passive tubular reabsorption, and secretion), diffusion characteristics may provide useful insight into the functional consequences of different renal diseases. DW MR imaging has been used to examine transplanted kidneys in an animal study (8). In the experimental transplant rejection model, ADC values in the cortex and medulla decreased significantly, suggesting the potential of this method in monitoring early graft rejection.
Blood oxygenation level–dependent (BOLD) MR imaging can be used to noninvasively assess the intrarenal oxygen content in native kidneys, provided all other factors—such as shimming status or measurement parameters—are kept constant (9–11). The relaxation rate R2* acquired with BOLD MR imaging is related to the tissue content of paramagnetic deoxyhemoglobin, which in turn is negatively related to the partial pressure of oxygen in blood (10).
Initial experience with BOLD MR imaging in patients with transplanted kidneys has been reported, demonstrating a significantly lower R2* in eight patients with acute allograft rejection as compared with six patients with normally functioning kidney transplants and six patients with acute tubular necrosis (12). However, information regarding reproducibility of results and comparisons with native kidneys was not provided.
BOLD MR imaging cannot be used to distinguish between changes in oxygenation attributed to alterations in perfusion and those attributed to alterations in oxygen consumption. Accordingly, simultaneous measurement of the perfusion fraction with DW MR imaging and of oxygenation with BOLD MR imaging may provide additional information.
The purpose of our study was to prospectively evaluate feasibility and reproducibility of DW and BOLD MR imaging in patients with renal allografts, as compared with these features in healthy volunteers with native kidneys.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
|
All subjects were allowed to consume 1 L of a nonalcoholic beverage 1–3 hours before MR imaging. Blood samples were obtained to measure the serum creatinine concentration immediately after MR imaging.
MR Imaging
MR imaging was performed with a 1.5-T imager (Sonata; Siemens, Erlangen, Germany) and a six-channel body coil. For morphologic evaluation, studies were acquired with a coronal T2-weighted half-Fourier single-shot fast spin-echo sequence and transverse and coronal T1-weighted fast low-angle shot gradient-echo sequences.
For functional evaluation, coronal multisection echo-planar DW MR imaging was performed with the following parameters: 21 sections (section thickness, 5 mm; intersection gap, 1 mm); field of view, 400 x 400 mm; matrix, 128 x 128; six signals acquired; bandwidth, 1500 Hz per pixel; and partial Fourier factor, 6/8. The following 10 diffusion gradient b values were used: 0, 10, 20, 40, 60, 150, 300, 500, 700, and 900 sec/mm2. The gradients were applied in three orthogonal directions and subsequently averaged to minimize the effects of diffusion anisotropy. A parallel imaging technique (modified sensitivity encoding) with a reduction factor of two was applied. Respiratory triggering was used with a minimum repetition time of 3200 msec and an echo time of 71 msec to reduce motion artifacts. Section positioning was identical to that used with the coronal T1-weighted sequence. Minimum acquisition time was 9 minutes 20 seconds.
A multiple gradient-recalled-echo sequence was used for BOLD MR imaging (9). Four or five coronal sections were acquired with a 5-mm section thickness and a 1-mm intersection gap. Field of view and matrix were 400 x 400 mm and 256 x 256, respectively; one signal was acquired. Other parameters were as follows: repetition time msec/echo time msec, 65/6–52; interecho spacing time, 4.2 msec; flip angle, 30°; and bandwidth, 325 Hz per pixel. Twelve T2*-weighted images corresponding to 12 different gradient echoes were acquired for each section within one 17-second breath hold. All examinations were completed successfully, and morphologic evaluation revealed no major complications.
Functional Evaluation
All participants' studies were included in the analysis. DW and BOLD MR imaging findings were analyzed independently by different observers (S.S., D.Z., H.C.T.) who were blinded to serum creatinine levels to eliminate bias.
DW MR imaging.—Diffusion values were calculated on a pixel-by-pixel basis in two ways. A total ADC value (ADCtot) was calculated with a weighted linear fit of ln(Si) with the following equation:
 | (1) |
To separate diffusion and microperfusion contributions to signal decay (5), biexponential fitting was performed with a Levenberg-Marquardt fitting algorithm, as follows:
 | (2) |
BOLD MR imaging.—R2* (1/T2*) maps were calculated on a pixel-by-pixel basis by fitting a weighted linear function through the logarithms of the signal intensities ln(Si) versus their respective echo times.
Image Analysis
Ellipsoid regions of interest (ROIs) were placed in the upper pole, middle pole, and lower pole of the cortex and medulla on several sections that covered large parts of the kidney for both DW MR imaging (mean number of sections, 6.6 ± 1.2; mean number of ROIs, 19.6 ± 3.6 for both the cortex and the medulla; mean individual ROI, 0.33 cm3 ± 0.13 for the cortex and 0.30 cm3 ± 0.12 for the medulla) and BOLD MR imaging (mean number of sections, 4.9 ± 0.4; mean number of ROIs, 14.4 ± 1.1 for both the cortex and the medulla; mean individual ROI, 0.16 cm3 ± 0.04 for the cortex and 0.18 cm3 ± 0.03 for the medulla). For DW MR imaging, two readers (H.C.T., 9 and 4 years of experience with renal MR imaging and DW MR imaging, respectively; D.Z., 1 year of experience with DW MR imaging) working in consensus simultaneously placed ROIs on coronal T1-weighted images and on corresponding DW images obtained with a b value of 0 sec/mm2. ROIs were placed on BOLD MR images obtained with a 6-msec echo time (S.S., 3 years of experience with BOLD MR imaging). Single total ROIs were created separately for the cortex and the medulla by merging all individual ROIs, yielding two ROIs for each subject (one ROI for the cortex and one for the medulla). The same procedure was used to evaluate native kidneys; in addition, ROIs in the right and left kidneys were merged for each subject separately for the cortex and the medulla after excluding significant right-left differences.
Statistical Analysis
Sample size for both the patient group and the volunteer group was estimated with a power analysis based on previous DW MR imaging findings in native kidneys (13); we assumed a similar standard deviation and required the corticomedullary difference to be detected with a significance level of .05 and 80% statistical power. A general linear model for repeated measurements (repeated-measures analysis of variance) was used with Bonferroni correction for multiple comparisons for statistical analysis of the results. In addition, 95% confidence intervals were calculated.
To test reproducibility in the seven patients who underwent repeat examinations, residual coefficients of variation between and within subjects were calculated as the square root of the residual mean square; these values are given as the percentage of the mean value (14). Pearson linear regression analysis was used to compare functional MR parameters between the cortex and the medulla, functional MR parameters with serum creatinine levels, and BOLD and DW MR imaging results. For all statistical tests, a P value of less than .05 was assumed to indicate statistical significance. Statistical analysis was performed with SPSS, version 12.0 (SPSS, Chicago, Ill), and Excel 2002 (Microsoft, Redmond, Wash) software.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
|
|
|
|
Reproducibility of DW MR Imaging Results
In patients with a renal allograft who underwent repeat MR imaging (Table 3), the within-subject ADCtot and ADCD were highly reproducible in the cortex and the medulla, with a coefficient of variation within subjects of less than 3.2%. The coefficient of variation within subjects for perfusion fraction was 8.6% in the cortex and 15.1% in the medulla. The variance between subjects was greater than the variance within subjects for all diffusion parameters; however, variance between subjects was still low for ADCtot and ADCD (
4.8%), whereas it was greater for the perfusion fraction (Fig 4).
|
|
|
5.2%, Table 3) and coefficient of variation between subjects was low in the cortex (4.1%) but much higher in the medulla (15.0%). In the repeated examinations, R2* correlation was strong in the medulla (r > 0.97) but not in the cortex (r = –0.14) because of the low cortical coefficient of variation between subjects (Fig 4).
Comparison of DW and BOLD MR Imaging
In transplanted kidneys, R2* correlated negatively with ADCtot (r = –0.66, P < .01; Fig 6) and ADCD (r = –0.61, P < .05) in the medulla. Other correlations between R2* and diffusion values were not significant. In native kidneys, R2* correlated negatively with ADCtot (r = –0.69, P < .01) and ADCD (r = –0.58, P < .05) in the cortex and with the perfusion fraction in the cortex (r = –0.64, P < .01) and the medulla (r = –0.61, P < .05).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
The ADCs of the healthy native kidneys were significantly higher in the cortex than in the medulla; this finding corroborated the findings of previous studies (8,13,15). In contrast, studies in which only low b values were used revealed that ADC values were higher in the medulla than in the cortex (16,17). These differences in ADC values may be attributed to the choice of b value ranges with varying contributions of diffusion and perfusion. This effect was largely eliminated in our study because we calculated true diffusion ADCD.
To our knowledge, ADC values in patients with transplanted kidneys have not been reported previously. In contrast to findings in native kidneys, ADC values were virtually identical in the cortex and the medulla of transplanted kidneys. Yang et al (8) used DW MR imaging to assess transplanted kidneys in rats. In agreement with our findings in human kidneys, the ADC values in the native rat kidneys were higher in the cortex than in the medulla, and these corticomedullary differences were smaller in transplanted rat kidneys. In contrast to our findings, the ADC values in the rats with renal allografts were lower than the ADC values in the rats with native kidneys. Renal allografts in the rats were investigated during the first 4 days after transplantation, whereas our measurements were obtained more than 100 days after transplantation; thus, a direct comparison of the study findings was impossible.
DW MR imaging simultaneously provides information on diffusion and either microcirculation or perfusion (5). The value of this separation between diffusion and microcirculation has been discussed (5,18,19), and only a few researchers have distinguished diffusion from perfusion in the kidneys (7,13,20,21). The perfusion fraction reflects microcirculation of blood and movement in predefined structures, such as tubular flow and glomerular filtration in the kidneys. Corresponding to the substantially higher regional blood flow in the renal cortex, in our study, the perfusion fraction obtained in the cortex was significantly higher than that obtained in the medulla in native kidneys. However, the corticomedullary difference was relatively small and was not significantly different in the transplanted kidneys. This suggests that perfusion fraction is influenced predominantly by factors other than blood perfusion, such as tubular flow.
BOLD MR imaging was found to be more sensitive when it was used to measure the oxygenation of red blood cells in the relatively hypoxic renal medulla than when it was used in the renal cortex, which has a higher partial pressure of oxygen in blood (10,22). This is in line with our findings of a low coefficient of variation within subjects and a high coefficient of variation between subjects for R2* in the medulla and thus indicates a sufficient sensitivity of medullary R2* for reliable assessment of oxygenation changes. In contrast, the similarly low values for coefficient of variation both within subjects and between subjects in the cortex suggest a relative insensitivity in the evaluation of oxygenation changes in the cortex.
In accordance with the findings of previous studies (10), R2* values were higher in the medulla than in the cortex, indicating a lower medullary oxygen content. Medullary R2* values and corticomedullary R2* differences were lower in transplanted kidneys than in native kidneys, suggesting a relatively increased oxygen content in the medulla of transplanted kidneys. The increased medullary oxygen content in patients with renal transplants may reflect the known reduced tubular fractional reabsorption of sodium (23,24,25). Alternatively, as the BOLD signal intensity is influenced by blood flow, volume, and oxygen concentration (26), the increased medullary oxygen content in transplanted kidneys may be attributed to increased blood flow due to allograft denervation.
The findings of animal studies disagree as to whether acute renal denervation increases blood flow in both the cortex and the medulla (27) or in the cortex alone (28,29). However, in contrast to increased medullary oxygenation, no increased perfusion fraction was observed in the cortex or medulla of patients with a renal transplant. In this respect, it is interesting that the significant correlation between R2* and perfusion fraction in the cortex and the medulla of native kidneys disappeared in transplanted kidneys. The finding of a similar perfusion fraction in transplanted and native kidneys may be partly related to the counteracting effects of the concomitant therapy in patients with a renal transplant, specifically cyclosporin A and medication that reduces angiotensin II effects, both of which diminish renal blood flow (30,31). The apparent counteracting effects of renal denervation and medication in patients with renal transplants preclude an unambiguous interpretation of the present results. Medullary oxygenation was higher in transplanted kidneys than in native kidneys, despite an unchanged perfusion fraction. This finding, as well as the correlation between oxygenation and perfusion fraction seen in native kidneys but not in transplanted kidneys, clearly indicates that BOLD MR imaging and DW MR imaging provide complementary information.
The significant correlation in our study between serum creatinine concentrations and some of the DW MR imaging parameters, such as perfusion fraction and ADCtot in the cortex, is an interesting finding that needs to be confirmed in a subsequent study involving a cohort of patients with renal transplants. These results are supported by the findings of previous studies and provide evidence of lower ADC values in the native kidneys of patients with increased serum creatinine concentrations (13,16,32). Thus, ADC values appear to be determined, at least in part, by using the glomerular filtration rate. There was no linear relationship in our study between perfusion fraction and serum creatinine level in native kidneys with serum creatinine concentrations of less than 100 µmol/L. This might have been due to either the narrow range of serum creatinine concentrations in healthy native kidneys or the presence of a serum creatinine concentration threshold above which the perfusion fraction decreases.
The unexpected finding of a low serum creatinine concentration in one patient (patient 8) together with low ADC values and high R2* in the medulla may have been due to underlying heterozygous hemoglobin E thalassemia, which is associated with oxidative instability and decreased oxygen saturation (33–35). The resultant high levels of deoxygenated hemoglobin cause an increase in R2*. Cell swelling may occur as a result of relative medullary hypoxia and lead to lower ADC values. We believe our study is the first in which a patient with hemoglobin E thalassemia was examined with DW and BOLD MR imaging; therefore, these findings must be considered preliminary.
There were limitations to our study. First, histopathologic correlation was needed to better define stable kidney function and to more precisely understand and interpret the functional MR findings. Histopathologic correlation will also be needed in further studies investigating pathologic allografts. Second, the correlations between serum creatinine concentrations and DW MR imaging parameters and between BOLD and DW MR imaging parameters need to be confirmed in a subsequent cohort of renal transplant patients, possibly one that includes subjects with a wider serum creatinine concentration range. Third, determination of reproducibility was based on a relatively small number of repeated studies in only seven subjects. However, because of distinct results for coefficients of variation within and between subjects for ADCtot, ADCD, and R2*, with a low coefficient of variation within subjects and an almost identical or distinctively higher coefficient of variation between subjects, we consider this number sufficient to estimate the reproducibility. Fourth, coefficients of variation in perfusion fraction between subjects and within subjects were relatively high in our study; this may have been due to blood pulsation and probably could have been reduced with cardiac triggering.
In conclusion, DW and BOLD MR imaging yield information on kidney function, as exemplified here in renal transplant patients. Both techniques provide highly reproducible results in subjects with good renal allograft function and hold promise for noninvasive monitoring of complications after kidney transplantation. However, to evaluate the potential of these methods in the assessment of functional derangements in patients with renal allografts, larger-scale studies must be performed; ideally, these studies should be linked with histopathologic analysis, if possible.
|
ADVANCE IN KNOWLEDGE |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
|
FOOTNOTES |
|---|
Abbreviations: ADC = apparent diffusion coefficient • BOLD = blood oxygenation level dependent • DW = diffusion weighted • ROI = region of interest
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, H.C.T., P. Vermathen; 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, H.C.T., D.Z., S.S., U.E., P. Vermathen; clinical studies, H.C.T., D.Z., U.E., P. Vermathen; statistical analysis, S.S., U.E., F.D.K., L.H., C.B., P. Vermathen; and manuscript editing, H.C.T., U.E., L.H., P. Vock, C.B., P. Vermathen.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCE IN KNOWLEDGEReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. Han, W. Xiao, Y. Xu, J. Wu, Q. Wang, H. Wang, M. Zhang, and J. Chen The significance of BOLD MRI in differentiation between renal transplant rejection and acute tubular necrosis Nephrol. Dial. Transplant., August 1, 2008; 23(8): 2666 - 2672. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. J. Wang and B. M. Yeh Is Assessing Renal Oxygenation by Using Blood Oxygen Level-Dependent MR Imaging a Clinical Reality? Radiology, June 1, 2008; 247(3): 595 - 596. [Full Text] [PDF]
|
|
|
|
|
|
H. C. Thoeny, T. M. Kessler, S. Simon-Zoula, F. De Keyzer, M. Mohaupt, U. E. Studer, and P. Vermathen Renal Oxygenation Changes during Acute Unilateral Ureteral Obstruction: Assessment with Blood Oxygen Level-Dependent MR Imaging--Initial Experience Radiology, June 1, 2008; 247(3): 754 - 761. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Zhang, Y. Mazaheri Tehrani, L. Wang, N. M. Ishill, L. H. Schwartz, and H. Hricak Renal Masses: Characterization with Diffusion-weighted MR Imaging--A Preliminary Experience Radiology, May 1, 2008; 247(2): 458 - 464. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
= patient 1, 
= patient 2, 
= patient 3, 
= patient 4, gray diamond = patient 5, gray square = patient 6, gray triangle = patient 7, gray circle = patient 8, 
= patient 9, 
= patient 10, 
= patient 11, 
= patient 12, + = patient 13, x = patient 14, * = patient 15). In some patients measurements were obtained twice. For the regression analysis (dashed line), the mean value of repeated measurements was considered.
ADCtot (cortex-medulla)) (r = –0.78, P < .01) declined in transplanted kidneys, indicating that increasing serum creatinine concentrations are accompanied by reduced cortical diffusion and microperfusion. For regression analyses, only transplanted kidneys were considered. The circled point indicates datum obtained in the patient with hemoglobinopathy.